chore(main): release 0.9.0 (#603)

* fix: kv use tls explicit

* fix: token refresh locking

* remove logs

* compile fix

* compile fix

---------

Co-authored-by: Igor Loskutov <igor.loskutoff@gmail.com>

.github/workflows/db_migrations.yml

@@ -2,6 +2,8 @@ name: Test Database Migrations
 
 on:
   push:
+    branches:
+      - main
     paths:
       - "server/migrations/**"
       - "server/reflector/db/**"
@@ -17,6 +19,9 @@ on:
 jobs:
   test-migrations:
     runs-on: ubuntu-latest
+    concurrency:
+      group: db-ubuntu-latest-${{ github.ref }}
+      cancel-in-progress: true
     services:
       postgres:
         image: postgres:17

.github/workflows/test_next_server.yml

@@ -0,0 +1,45 @@
+name: Test Next Server
+
+on:
+  pull_request:
+    paths:
+      - "www/**"
+  push:
+    branches:
+      - main
+    paths:
+      - "www/**"
+
+jobs:
+  test-next-server:
+    runs-on: ubuntu-latest
+
+    defaults:
+      run:
+        working-directory: ./www
+
+    steps:
+    - uses: actions/checkout@v4
+
+    - name: Setup Node.js
+      uses: actions/setup-node@v4
+      with:
+        node-version: '20'
+
+    - name: Install pnpm
+      uses: pnpm/action-setup@v4
+      with:
+        version: 8
+
+    - name: Setup Node.js cache
+      uses: actions/setup-node@v4
+      with:
+        node-version: '20'
+        cache: 'pnpm'
+        cache-dependency-path: './www/pnpm-lock.yaml'
+
+    - name: Install dependencies
+      run: pnpm install
+
+    - name: Run tests
+      run: pnpm test

.github/workflows/test_server.yml

@@ -5,12 +5,17 @@ on:
     paths:
       - "server/**"
   push:
+    branches:
+      - main
     paths:
       - "server/**"
 
 jobs:
   pytest:
     runs-on: ubuntu-latest
+    concurrency:
+      group: pytest-${{ github.ref }}
+      cancel-in-progress: true
     services:
       redis:
         image: redis:6
@@ -30,6 +35,9 @@ jobs:
 
   docker-amd64:
     runs-on: linux-amd64
+    concurrency:
+      group: docker-amd64-${{ github.ref }}
+      cancel-in-progress: true
     steps:
       - uses: actions/checkout@v4
       - name: Set up Docker Buildx
@@ -45,6 +53,9 @@ jobs:
 
   docker-arm64:
     runs-on: linux-arm64
+    concurrency:
+      group: docker-arm64-${{ github.ref }}
+      cancel-in-progress: true
     steps:
       - uses: actions/checkout@v4
       - name: Set up Docker Buildx

.gitignore

@@ -14,4 +14,7 @@ data/
 www/REFACTOR.md
 www/reload-frontend
 server/test.sqlite
-CLAUDE.local.md
+CLAUDE.local.md
+www/.env.development
+www/.env.production
+.playwright-mcp

.gitleaksignore

@@ -0,0 +1 @@
+b9d891d3424f371642cb032ecfd0e2564470a72c:server/tests/test_transcripts_recording_deletion.py:generic-api-key:15

.pre-commit-config.yaml

@@ -27,3 +27,8 @@ repos:
         files: ^server/
       - id: ruff-format
         files: ^server/
+
+  - repo: https://github.com/gitleaks/gitleaks
+    rev: v8.28.0
+    hooks:
+      - id: gitleaks

CHANGELOG.md

@@ -1,5 +1,57 @@
 # Changelog
 
+## [0.9.0](https://github.com/Monadical-SAS/reflector/compare/v0.8.2...v0.9.0) (2025-09-06)
+
+
+### Features
+
+* frontend openapi react query ([#606](https://github.com/Monadical-SAS/reflector/issues/606)) ([c4d2825](https://github.com/Monadical-SAS/reflector/commit/c4d2825c81f81ad8835629fbf6ea8c7383f8c31b))
+
+
+### Bug Fixes
+
+* align whisper transcriber api with parakeet ([#602](https://github.com/Monadical-SAS/reflector/issues/602)) ([0663700](https://github.com/Monadical-SAS/reflector/commit/0663700a615a4af69a03c96c410f049e23ec9443))
+* kv use tls explicit ([#610](https://github.com/Monadical-SAS/reflector/issues/610)) ([08d88ec](https://github.com/Monadical-SAS/reflector/commit/08d88ec349f38b0d13e0fa4cb73486c8dfd31836))
+* source kind for file processing ([#601](https://github.com/Monadical-SAS/reflector/issues/601)) ([dc82f8b](https://github.com/Monadical-SAS/reflector/commit/dc82f8bb3bdf3ab3d4088e592a30fd63907319e1))
+* token refresh locking ([#613](https://github.com/Monadical-SAS/reflector/issues/613)) ([7f5a4c9](https://github.com/Monadical-SAS/reflector/commit/7f5a4c9ddc7fd098860c8bdda2ca3b57f63ded2f))
+
+## [0.8.2](https://github.com/Monadical-SAS/reflector/compare/v0.8.1...v0.8.2) (2025-08-29)
+
+
+### Bug Fixes
+
+* search-logspam ([#593](https://github.com/Monadical-SAS/reflector/issues/593)) ([695d1a9](https://github.com/Monadical-SAS/reflector/commit/695d1a957d4cd862753049f9beed88836cabd5ab))
+
+## [0.8.1](https://github.com/Monadical-SAS/reflector/compare/v0.8.0...v0.8.1) (2025-08-29)
+
+
+### Bug Fixes
+
+* make webhook secret/url allowing null ([#590](https://github.com/Monadical-SAS/reflector/issues/590)) ([84a3812](https://github.com/Monadical-SAS/reflector/commit/84a381220bc606231d08d6f71d4babc818fa3c75))
+
+## [0.8.0](https://github.com/Monadical-SAS/reflector/compare/v0.7.3...v0.8.0) (2025-08-29)
+
+
+### Features
+
+* **cleanup:** add automatic data retention for public instances ([#574](https://github.com/Monadical-SAS/reflector/issues/574)) ([6f0c7c1](https://github.com/Monadical-SAS/reflector/commit/6f0c7c1a5e751713366886c8e764c2009e12ba72))
+* **rooms:** add webhook for transcript completion ([#578](https://github.com/Monadical-SAS/reflector/issues/578)) ([88ed7cf](https://github.com/Monadical-SAS/reflector/commit/88ed7cfa7804794b9b54cad4c3facc8a98cf85fd))
+
+
+### Bug Fixes
+
+* file pipeline status reporting and websocket updates ([#589](https://github.com/Monadical-SAS/reflector/issues/589)) ([9dfd769](https://github.com/Monadical-SAS/reflector/commit/9dfd76996f851cc52be54feea078adbc0816dc57))
+* Igor/evaluation ([#575](https://github.com/Monadical-SAS/reflector/issues/575)) ([124ce03](https://github.com/Monadical-SAS/reflector/commit/124ce03bf86044c18313d27228a25da4bc20c9c5))
+* optimize parakeet transcription batching algorithm ([#577](https://github.com/Monadical-SAS/reflector/issues/577)) ([7030e0f](https://github.com/Monadical-SAS/reflector/commit/7030e0f23649a8cf6c1eb6d5889684a41ce849ec))
+
+## [0.7.3](https://github.com/Monadical-SAS/reflector/compare/v0.7.2...v0.7.3) (2025-08-22)
+
+
+### Bug Fixes
+
+* cleaned repo, and get git-leaks clean ([359280d](https://github.com/Monadical-SAS/reflector/commit/359280dd340433ba4402ed69034094884c825e67))
+* restore previous behavior on live pipeline + audio downscaler ([#561](https://github.com/Monadical-SAS/reflector/issues/561)) ([9265d20](https://github.com/Monadical-SAS/reflector/commit/9265d201b590d23c628c5f19251b70f473859043))
+
 ## [0.7.2](https://github.com/Monadical-SAS/reflector/compare/v0.7.1...v0.7.2) (2025-08-21)
 
 

README.md

@@ -1,43 +1,60 @@
 <div align="center">
+<img width="100" alt="image" src="https://github.com/user-attachments/assets/66fb367b-2c89-4516-9912-f47ac59c6a7f"/>
 
 # Reflector
 
-Reflector Audio Management and Analysis is a cutting-edge web application under development by Monadical. It utilizes AI to record meetings, providing a permanent record with transcripts, translations, and automated summaries.
+Reflector is an AI-powered audio transcription and meeting analysis platform that provides real-time transcription, speaker diarization, translation and summarization for audio content and live meetings. It works 100% with local models (whisper/parakeet, pyannote, seamless-m4t, and your local llm like phi-4).
 
-[![Tests](https://github.com/monadical-sas/reflector/actions/workflows/pytests.yml/badge.svg?branch=main&event=push)](https://github.com/monadical-sas/reflector/actions/workflows/pytests.yml)
+[![Tests](https://github.com/monadical-sas/reflector/actions/workflows/test_server.yml/badge.svg?branch=main&event=push)](https://github.com/monadical-sas/reflector/actions/workflows/test_server.yml)
 [![License: MIT](https://img.shields.io/badge/license-MIT-green.svg)](https://opensource.org/licenses/MIT)
 </div>
-
-## Screenshots
+</div>
 <table>
   <tr>
     <td>
-      <a href="https://github.com/user-attachments/assets/3a976930-56c1-47ef-8c76-55d3864309e3">
-        <img width="700" alt="image" src="https://github.com/user-attachments/assets/3a976930-56c1-47ef-8c76-55d3864309e3" />
+      <a href="https://github.com/user-attachments/assets/21f5597c-2930-4899-a154-f7bd61a59e97">
+        <img width="700" alt="image" src="https://github.com/user-attachments/assets/21f5597c-2930-4899-a154-f7bd61a59e97" />
       </a>
     </td>
     <td>
-      <a href="https://github.com/user-attachments/assets/bfe3bde3-08af-4426-a9a1-11ad5cd63b33">
-        <img width="700" alt="image" src="https://github.com/user-attachments/assets/bfe3bde3-08af-4426-a9a1-11ad5cd63b33" />
+      <a href="https://github.com/user-attachments/assets/f6b9399a-5e51-4bae-b807-59128d0a940c">
+        <img width="700" alt="image" src="https://github.com/user-attachments/assets/f6b9399a-5e51-4bae-b807-59128d0a940c" />
       </a>
     </td>
     <td>
-      <a href="https://github.com/user-attachments/assets/7b60c9d0-efe4-474f-a27b-ea13bd0fabdc">
-        <img width="700" alt="image" src="https://github.com/user-attachments/assets/7b60c9d0-efe4-474f-a27b-ea13bd0fabdc" />
+      <a href="https://github.com/user-attachments/assets/a42ce460-c1fd-4489-a995-270516193897">
+        <img width="700" alt="image" src="https://github.com/user-attachments/assets/a42ce460-c1fd-4489-a995-270516193897" />
+      </a>
+    </td>
+    <td>
+      <a href="https://github.com/user-attachments/assets/21929f6d-c309-42fe-9c11-f1299e50fbd4">
+        <img width="700" alt="image" src="https://github.com/user-attachments/assets/21929f6d-c309-42fe-9c11-f1299e50fbd4" />
       </a>
     </td>
   </tr>
 </table>
 
+## What is Reflector?
+
+Reflector is a web application that utilizes local models to process audio content, providing:
+
+- **Real-time Transcription**: Convert speech to text using [Whisper](https://github.com/openai/whisper) (multi-language) or [Parakeet](https://huggingface.co/nvidia/parakeet-tdt-0.6b-v2) (English) models
+- **Speaker Diarization**: Identify and label different speakers using [Pyannote](https://github.com/pyannote/pyannote-audio) 3.1
+- **Live Translation**: Translate audio content in real-time to many languages with [Facebook Seamless-M4T](https://github.com/facebookresearch/seamless_communication)
+- **Topic Detection & Summarization**: Extract key topics and generate concise summaries using LLMs
+- **Meeting Recording**: Create permanent records of meetings with searchable transcripts
+
+Currently we provide [modal.com](https://modal.com/) gpu template to deploy.
+
 ## Background
 
 The project architecture consists of three primary components:
 
-- **Front-End**: NextJS React project hosted on Vercel, located in `www/`.
 - **Back-End**: Python server that offers an API and data persistence, found in `server/`.
-- **GPU implementation**: Providing services such as speech-to-text transcription, topic generation, automated summaries, and translations. Most reliable option is Modal deployment
+- **Front-End**: NextJS React project hosted on Vercel, located in `www/`.
+- **GPU implementation**: Providing services such as speech-to-text transcription, topic generation, automated summaries, and translations.
 
-It also uses authentik for authentication if activated, and Vercel for deployment and configuration of the front-end.
+It also uses authentik for authentication if activated.
 
 ## Contribution Guidelines
 
@@ -72,6 +89,8 @@ Note: We currently do not have instructions for Windows users.
 
 ## Installation
 
+*Note: we're working toward better installation, theses instructions are not accurate for now*
+
 ### Frontend
 
 Start with `cd www`.

compose.yml

@@ -6,6 +6,7 @@ services:
       - 1250:1250
     volumes:
       - ./server/:/app/
+      - /app/.venv
     env_file:
       - ./server/.env
     environment:
@@ -16,6 +17,7 @@ services:
       context: server
     volumes:
       - ./server/:/app/
+      - /app/.venv
     env_file:
       - ./server/.env
     environment:
@@ -26,6 +28,7 @@ services:
       context: server
     volumes:
       - ./server/:/app/
+      - /app/.venv
     env_file:
       - ./server/.env
     environment:

server/docs/data_retention.md

@@ -0,0 +1,95 @@
+# Data Retention and Cleanup
+
+## Overview
+
+For public instances of Reflector, a data retention policy is automatically enforced to delete anonymous user data after a configurable period (default: 7 days). This ensures compliance with privacy expectations and prevents unbounded storage growth.
+
+## Configuration
+
+### Environment Variables
+
+- `PUBLIC_MODE` (bool): Must be set to `true` to enable automatic cleanup
+- `PUBLIC_DATA_RETENTION_DAYS` (int): Number of days to retain anonymous data (default: 7)
+
+### What Gets Deleted
+
+When data reaches the retention period, the following items are automatically removed:
+
+1. **Transcripts** from anonymous users (where `user_id` is NULL):
+   - Database records
+   - Local files (audio.wav, audio.mp3, audio.json waveform)
+   - Storage files (cloud storage if configured)
+
+## Automatic Cleanup
+
+### Celery Beat Schedule
+
+When `PUBLIC_MODE=true`, a Celery beat task runs daily at 3 AM to clean up old data:
+
+```python
+# Automatically scheduled when PUBLIC_MODE=true
+"cleanup_old_public_data": {
+    "task": "reflector.worker.cleanup.cleanup_old_public_data",
+    "schedule": crontab(hour=3, minute=0),  # Daily at 3 AM
+}
+```
+
+### Running the Worker
+
+Ensure both Celery worker and beat scheduler are running:
+
+```bash
+# Start Celery worker
+uv run celery -A reflector.worker.app worker --loglevel=info
+
+# Start Celery beat scheduler (in another terminal)
+uv run celery -A reflector.worker.app beat
+```
+
+## Manual Cleanup
+
+For testing or manual intervention, use the cleanup tool:
+
+```bash
+# Delete data older than 7 days (default)
+uv run python -m reflector.tools.cleanup_old_data
+
+# Delete data older than 30 days
+uv run python -m reflector.tools.cleanup_old_data --days 30
+```
+
+Note: The manual tool uses the same implementation as the Celery worker task to ensure consistency.
+
+## Important Notes
+
+1. **User Data Deletion**: Only anonymous data (where `user_id` is NULL) is deleted. Authenticated user data is preserved.
+
+2. **Storage Cleanup**: The system properly cleans up both local files and cloud storage when configured.
+
+3. **Error Handling**: If individual deletions fail, the cleanup continues and logs errors. Failed deletions are reported in the task output.
+
+4. **Public Instance Only**: The automatic cleanup task only runs when `PUBLIC_MODE=true` to prevent accidental data loss in private deployments.
+
+## Testing
+
+Run the cleanup tests:
+
+```bash
+uv run pytest tests/test_cleanup.py -v
+```
+
+## Monitoring
+
+Check Celery logs for cleanup task execution:
+
+```bash
+# Look for cleanup task logs
+grep "cleanup_old_public_data" celery.log
+grep "Starting cleanup of old public data" celery.log
+```
+
+Task statistics are logged after each run:
+- Number of transcripts deleted
+- Number of meetings deleted
+- Number of orphaned recordings deleted
+- Any errors encountered

server/docs/gpu/api-transcription.md

@@ -0,0 +1,194 @@
+## Reflector GPU Transcription API (Specification)
+
+This document defines the Reflector GPU transcription API that all implementations must adhere to. Current implementations include NVIDIA Parakeet (NeMo) and Whisper (faster-whisper), both deployed on Modal.com. The API surface and response shapes are OpenAI/Whisper-compatible, so clients can switch implementations by changing only the base URL.
+
+### Base URL and Authentication
+
+- Example base URLs (Modal web endpoints):
+
+  - Parakeet: `https://<account>--reflector-transcriber-parakeet-web.modal.run`
+  - Whisper: `https://<account>--reflector-transcriber-web.modal.run`
+
+- All endpoints are served under `/v1` and require a Bearer token:
+
+```
+Authorization: Bearer <REFLECTOR_GPU_APIKEY>
+```
+
+Note: To switch implementations, deploy the desired variant and point `TRANSCRIPT_URL` to its base URL. The API is identical.
+
+### Supported file types
+
+`mp3, mp4, mpeg, mpga, m4a, wav, webm`
+
+### Models and languages
+
+- Parakeet (NVIDIA NeMo): default `nvidia/parakeet-tdt-0.6b-v2`
+  - Language support: only `en`. Other languages return HTTP 400.
+- Whisper (faster-whisper): default `large-v2` (or deployment-specific)
+  - Language support: multilingual (per Whisper model capabilities).
+
+Note: The `model` parameter is accepted by all implementations for interface parity. Some backends may treat it as informational.
+
+### Endpoints
+
+#### POST /v1/audio/transcriptions
+
+Transcribe one or more uploaded audio files.
+
+Request: multipart/form-data
+
+- `file` (File) — optional. Single file to transcribe.
+- `files` (File[]) — optional. One or more files to transcribe.
+- `model` (string) — optional. Defaults to the implementation-specific model (see above).
+- `language` (string) — optional, defaults to `en`.
+  - Parakeet: only `en` is accepted; other values return HTTP 400
+  - Whisper: model-dependent; typically multilingual
+- `batch` (boolean) — optional, defaults to `false`.
+
+Notes:
+
+- Provide either `file` or `files`, not both. If neither is provided, HTTP 400.
+- `batch` requires `files`; using `batch=true` without `files` returns HTTP 400.
+- Response shape for multiple files is the same regardless of `batch`.
+- Files sent to this endpoint are processed in a single pass (no VAD/chunking). This is intended for short clips (roughly ≤ 30s; depends on GPU memory/model). For longer audio, prefer `/v1/audio/transcriptions-from-url` which supports VAD-based chunking.
+
+Responses
+
+Single file response:
+
+```json
+{
+  "text": "transcribed text",
+  "words": [
+    { "word": "hello", "start": 0.0, "end": 0.5 },
+    { "word": "world", "start": 0.5, "end": 1.0 }
+  ],
+  "filename": "audio.mp3"
+}
+```
+
+Multiple files response:
+
+```json
+{
+  "results": [
+    {"filename": "a1.mp3", "text": "...", "words": [...]},
+    {"filename": "a2.mp3", "text": "...", "words": [...]}]
+}
+```
+
+Notes:
+
+- Word objects always include keys: `word`, `start`, `end`.
+- Some implementations may include a trailing space in `word` to match Whisper tokenization behavior; clients should trim if needed.
+
+Example curl (single file):
+
+```bash
+curl -X POST \
+  -H "Authorization: Bearer $REFLECTOR_GPU_APIKEY" \
+  -F "file=@/path/to/audio.mp3" \
+  -F "language=en" \
+  "$BASE_URL/v1/audio/transcriptions"
+```
+
+Example curl (multiple files, batch):
+
+```bash
+curl -X POST \
+  -H "Authorization: Bearer $REFLECTOR_GPU_APIKEY" \
+  -F "files=@/path/a1.mp3" -F "files=@/path/a2.mp3" \
+  -F "batch=true" -F "language=en" \
+  "$BASE_URL/v1/audio/transcriptions"
+```
+
+#### POST /v1/audio/transcriptions-from-url
+
+Transcribe a single remote audio file by URL.
+
+Request: application/json
+
+Body parameters:
+
+- `audio_file_url` (string) — required. URL of the audio file to transcribe.
+- `model` (string) — optional. Defaults to the implementation-specific model (see above).
+- `language` (string) — optional, defaults to `en`. Parakeet only accepts `en`.
+- `timestamp_offset` (number) — optional, defaults to `0.0`. Added to each word's `start`/`end` in the response.
+
+```json
+{
+  "audio_file_url": "https://example.com/audio.mp3",
+  "model": "nvidia/parakeet-tdt-0.6b-v2",
+  "language": "en",
+  "timestamp_offset": 0.0
+}
+```
+
+Response:
+
+```json
+{
+  "text": "transcribed text",
+  "words": [
+    { "word": "hello", "start": 10.0, "end": 10.5 },
+    { "word": "world", "start": 10.5, "end": 11.0 }
+  ]
+}
+```
+
+Notes:
+
+- `timestamp_offset` is added to each word’s `start`/`end` in the response.
+- Implementations may perform VAD-based chunking and batching for long-form audio; word timings are adjusted accordingly.
+
+Example curl:
+
+```bash
+curl -X POST \
+  -H "Authorization: Bearer $REFLECTOR_GPU_APIKEY" \
+  -H "Content-Type: application/json" \
+  -d '{
+        "audio_file_url": "https://example.com/audio.mp3",
+        "language": "en",
+        "timestamp_offset": 0
+      }' \
+  "$BASE_URL/v1/audio/transcriptions-from-url"
+```
+
+### Error handling
+
+- 400 Bad Request
+  - Parakeet: `language` other than `en`
+  - Missing required parameters (`file`/`files` for upload; `audio_file_url` for URL endpoint)
+  - Unsupported file extension
+- 401 Unauthorized
+  - Missing or invalid Bearer token
+- 404 Not Found
+  - `audio_file_url` does not exist
+
+### Implementation details
+
+- GPUs: A10G for small-file/live, L40S for large-file URL transcription (subject to deployment)
+- VAD chunking and segment batching; word timings adjusted and overlapping ends constrained
+- Pads very short segments (< 0.5s) to avoid model crashes on some backends
+
+### Server configuration (Reflector API)
+
+Set the Reflector server to use the Modal backend and point `TRANSCRIPT_URL` to your chosen deployment:
+
+```
+TRANSCRIPT_BACKEND=modal
+TRANSCRIPT_URL=https://<account>--reflector-transcriber-parakeet-web.modal.run
+TRANSCRIPT_MODAL_API_KEY=<REFLECTOR_GPU_APIKEY>
+```
+
+### Conformance tests
+
+Use the pytest-based conformance tests to validate any new implementation (including self-hosted) against this spec:
+
+```
+TRANSCRIPT_URL=https://<your-deployment-base> \
+TRANSCRIPT_MODAL_API_KEY=your-api-key \
+uv run -m pytest -m gpu_modal --no-cov server/tests/test_gpu_modal_transcript.py
+```

server/docs/webhook.md

@@ -0,0 +1,212 @@
+# Reflector Webhook Documentation
+
+## Overview
+
+Reflector supports webhook notifications to notify external systems when transcript processing is completed. Webhooks can be configured per room and are triggered automatically after a transcript is successfully processed.
+
+## Configuration
+
+Webhooks are configured at the room level with two fields:
+- `webhook_url`: The HTTPS endpoint to receive webhook notifications
+- `webhook_secret`: Optional secret key for HMAC signature verification (auto-generated if not provided)
+
+## Events
+
+### `transcript.completed`
+
+Triggered when a transcript has been fully processed, including transcription, diarization, summarization, and topic detection.
+
+### `test`
+
+A test event that can be triggered manually to verify webhook configuration.
+
+## Webhook Request Format
+
+### Headers
+
+All webhook requests include the following headers:
+
+| Header | Description | Example |
+|--------|-------------|---------|
+| `Content-Type` | Always `application/json` | `application/json` |
+| `User-Agent` | Identifies Reflector as the source | `Reflector-Webhook/1.0` |
+| `X-Webhook-Event` | The event type | `transcript.completed` or `test` |
+| `X-Webhook-Retry` | Current retry attempt number | `0`, `1`, `2`... |
+| `X-Webhook-Signature` | HMAC signature (if secret configured) | `t=1735306800,v1=abc123...` |
+
+### Signature Verification
+
+If a webhook secret is configured, Reflector includes an HMAC-SHA256 signature in the `X-Webhook-Signature` header to verify the webhook authenticity.
+
+The signature format is: `t={timestamp},v1={signature}`
+
+To verify the signature:
+1. Extract the timestamp and signature from the header
+2. Create the signed payload: `{timestamp}.{request_body}`
+3. Compute HMAC-SHA256 of the signed payload using your webhook secret
+4. Compare the computed signature with the received signature
+
+Example verification (Python):
+```python
+import hmac
+import hashlib
+
+def verify_webhook_signature(payload: bytes, signature_header: str, secret: str) -> bool:
+    # Parse header: "t=1735306800,v1=abc123..."
+    parts = dict(part.split("=") for part in signature_header.split(","))
+    timestamp = parts["t"]
+    received_signature = parts["v1"]
+
+    # Create signed payload
+    signed_payload = f"{timestamp}.{payload.decode('utf-8')}"
+
+    # Compute expected signature
+    expected_signature = hmac.new(
+        secret.encode("utf-8"),
+        signed_payload.encode("utf-8"),
+        hashlib.sha256
+    ).hexdigest()
+
+    # Compare signatures
+    return hmac.compare_digest(expected_signature, received_signature)
+```
+
+## Event Payloads
+
+### `transcript.completed` Event
+
+This event includes a convenient URL for accessing the transcript:
+- `frontend_url`: Direct link to view the transcript in the web interface
+
+```json
+{
+  "event": "transcript.completed",
+  "event_id": "transcript.completed-abc-123-def-456",
+  "timestamp": "2025-08-27T12:34:56.789012Z",
+  "transcript": {
+    "id": "abc-123-def-456",
+    "room_id": "room-789",
+    "created_at": "2025-08-27T12:00:00Z",
+    "duration": 1800.5,
+    "title": "Q3 Product Planning Meeting",
+    "short_summary": "Team discussed Q3 product roadmap, prioritizing mobile app features and API improvements.",
+    "long_summary": "The product team met to finalize the Q3 roadmap. Key decisions included...",
+    "webvtt": "WEBVTT\n\n00:00:00.000 --> 00:00:05.000\n<v Speaker 1>Welcome everyone to today's meeting...",
+    "topics": [
+      {
+        "title": "Introduction and Agenda",
+        "summary": "Meeting kickoff with agenda review",
+        "timestamp": 0.0,
+        "duration": 120.0,
+        "webvtt": "WEBVTT\n\n00:00:00.000 --> 00:00:05.000\n<v Speaker 1>Welcome everyone..."
+      },
+      {
+        "title": "Mobile App Features Discussion",
+        "summary": "Team reviewed proposed mobile app features for Q3",
+        "timestamp": 120.0,
+        "duration": 600.0,
+        "webvtt": "WEBVTT\n\n00:02:00.000 --> 00:02:10.000\n<v Speaker 2>Let's talk about the mobile app..."
+      }
+    ],
+    "participants": [
+      {
+        "id": "participant-1",
+        "name": "John Doe",
+        "speaker": "Speaker 1"
+      },
+      {
+        "id": "participant-2",
+        "name": "Jane Smith",
+        "speaker": "Speaker 2"
+      }
+    ],
+    "source_language": "en",
+    "target_language": "en",
+    "status": "completed",
+    "frontend_url": "https://app.reflector.com/transcripts/abc-123-def-456"
+  },
+  "room": {
+    "id": "room-789",
+    "name": "Product Team Room"
+  }
+}
+```
+
+### `test` Event
+
+```json
+{
+  "event": "test",
+  "event_id": "test.2025-08-27T12:34:56.789012Z",
+  "timestamp": "2025-08-27T12:34:56.789012Z",
+  "message": "This is a test webhook from Reflector",
+  "room": {
+    "id": "room-789",
+    "name": "Product Team Room"
+  }
+}
+```
+
+## Retry Policy
+
+Webhooks are delivered with automatic retry logic to handle transient failures. When a webhook delivery fails due to server errors or network issues, Reflector will automatically retry the delivery multiple times over an extended period.
+
+### Retry Mechanism
+
+Reflector implements an exponential backoff strategy for webhook retries:
+
+- **Initial retry delay**: 60 seconds after the first failure
+- **Exponential backoff**: Each subsequent retry waits approximately twice as long as the previous one
+- **Maximum retry interval**: 1 hour (backoff is capped at this duration)
+- **Maximum retry attempts**: 30 attempts total
+- **Total retry duration**: Retries continue for approximately 24 hours
+
+### How Retries Work
+
+When a webhook fails, Reflector will:
+1. Wait 60 seconds, then retry (attempt #1)
+2. If it fails again, wait ~2 minutes, then retry (attempt #2)
+3. Continue doubling the wait time up to a maximum of 1 hour between attempts
+4. Keep retrying at 1-hour intervals until successful or 30 attempts are exhausted
+
+The `X-Webhook-Retry` header indicates the current retry attempt number (0 for the initial attempt, 1 for first retry, etc.), allowing your endpoint to track retry attempts.
+
+### Retry Behavior by HTTP Status Code
+
+| Status Code | Behavior |
+|-------------|----------|
+| 2xx (Success) | No retry, webhook marked as delivered |
+| 4xx (Client Error) | No retry, request is considered permanently failed |
+| 5xx (Server Error) | Automatic retry with exponential backoff |
+| Network/Timeout Error | Automatic retry with exponential backoff |
+
+**Important Notes:**
+- Webhooks timeout after 30 seconds. If your endpoint takes longer to respond, it will be considered a timeout error and retried.
+- During the retry period (~24 hours), you may receive the same webhook multiple times if your endpoint experiences intermittent failures.
+- There is no mechanism to manually retry failed webhooks after the retry period expires.
+
+## Testing Webhooks
+
+You can test your webhook configuration before processing transcripts:
+
+```http
+POST /v1/rooms/{room_id}/webhook/test
+```
+
+Response:
+```json
+{
+  "success": true,
+  "status_code": 200,
+  "message": "Webhook test successful",
+  "response_preview": "OK"
+}
+```
+
+Or in case of failure:
+```json
+{
+  "success": false,
+  "error": "Webhook request timed out (10 seconds)"
+}
+```

server/gpu/modal_deployments/reflector_transcriber.py

@@ -1,41 +1,78 @@
 import os
-import tempfile
+import sys
 import threading
+import uuid
+from typing import Generator, Mapping, NamedTuple, NewType, TypedDict
+from urllib.parse import urlparse
 
 import modal
-from pydantic import BaseModel
-
-MODELS_DIR = "/models"
 
 MODEL_NAME = "large-v2"
 MODEL_COMPUTE_TYPE: str = "float16"
 MODEL_NUM_WORKERS: int = 1
-
 MINUTES = 60  # seconds
+SAMPLERATE = 16000
+UPLOADS_PATH = "/uploads"
+CACHE_PATH = "/models"
+SUPPORTED_FILE_EXTENSIONS = ["mp3", "mp4", "mpeg", "mpga", "m4a", "wav", "webm"]
+VAD_CONFIG = {
+    "batch_max_duration": 30.0,
+    "silence_padding": 0.5,
+    "window_size": 512,
+}
 
-volume = modal.Volume.from_name("models", create_if_missing=True)
+
+WhisperUniqFilename = NewType("WhisperUniqFilename", str)
+AudioFileExtension = NewType("AudioFileExtension", str)
 
 app = modal.App("reflector-transcriber")
 
+model_cache = modal.Volume.from_name("models", create_if_missing=True)
+upload_volume = modal.Volume.from_name("whisper-uploads", create_if_missing=True)
+
+
+class TimeSegment(NamedTuple):
+    """Represents a time segment with start and end times."""
+
+    start: float
+    end: float
+
+
+class AudioSegment(NamedTuple):
+    """Represents an audio segment with timing and audio data."""
+
+    start: float
+    end: float
+    audio: any
+
+
+class TranscriptResult(NamedTuple):
+    """Represents a transcription result with text and word timings."""
+
+    text: str
+    words: list["WordTiming"]
+
+
+class WordTiming(TypedDict):
+    """Represents a word with its timing information."""
+
+    word: str
+    start: float
+    end: float
+
 
 def download_model():
     from faster_whisper import download_model
 
-    volume.reload()
+    model_cache.reload()
 
-    download_model(MODEL_NAME, cache_dir=MODELS_DIR)
+    download_model(MODEL_NAME, cache_dir=CACHE_PATH)
 
-    volume.commit()
+    model_cache.commit()
 
 
 image = (
     modal.Image.debian_slim(python_version="3.12")
-    .pip_install(
-        "huggingface_hub==0.27.1",
-        "hf-transfer==0.1.9",
-        "torch==2.5.1",
-        "faster-whisper==1.1.1",
-    )
     .env(
         {
             "HF_HUB_ENABLE_HF_TRANSFER": "1",
@@ -45,19 +82,98 @@ image = (
             ),
         }
     )
-    .run_function(download_model, volumes={MODELS_DIR: volume})
+    .apt_install("ffmpeg")
+    .pip_install(
+        "huggingface_hub==0.27.1",
+        "hf-transfer==0.1.9",
+        "torch==2.5.1",
+        "faster-whisper==1.1.1",
+        "fastapi==0.115.12",
+        "requests",
+        "librosa==0.10.1",
+        "numpy<2",
+        "silero-vad==5.1.0",
+    )
+    .run_function(download_model, volumes={CACHE_PATH: model_cache})
 )
 
 
+def detect_audio_format(url: str, headers: Mapping[str, str]) -> AudioFileExtension:
+    parsed_url = urlparse(url)
+    url_path = parsed_url.path
+
+    for ext in SUPPORTED_FILE_EXTENSIONS:
+        if url_path.lower().endswith(f".{ext}"):
+            return AudioFileExtension(ext)
+
+    content_type = headers.get("content-type", "").lower()
+    if "audio/mpeg" in content_type or "audio/mp3" in content_type:
+        return AudioFileExtension("mp3")
+    if "audio/wav" in content_type:
+        return AudioFileExtension("wav")
+    if "audio/mp4" in content_type:
+        return AudioFileExtension("mp4")
+
+    raise ValueError(
+        f"Unsupported audio format for URL: {url}. "
+        f"Supported extensions: {', '.join(SUPPORTED_FILE_EXTENSIONS)}"
+    )
+
+
+def download_audio_to_volume(
+    audio_file_url: str,
+) -> tuple[WhisperUniqFilename, AudioFileExtension]:
+    import requests
+    from fastapi import HTTPException
+
+    response = requests.head(audio_file_url, allow_redirects=True)
+    if response.status_code == 404:
+        raise HTTPException(status_code=404, detail="Audio file not found")
+
+    response = requests.get(audio_file_url, allow_redirects=True)
+    response.raise_for_status()
+
+    audio_suffix = detect_audio_format(audio_file_url, response.headers)
+    unique_filename = WhisperUniqFilename(f"{uuid.uuid4()}.{audio_suffix}")
+    file_path = f"{UPLOADS_PATH}/{unique_filename}"
+
+    with open(file_path, "wb") as f:
+        f.write(response.content)
+
+    upload_volume.commit()
+    return unique_filename, audio_suffix
+
+
+def pad_audio(audio_array, sample_rate: int = SAMPLERATE):
+    """Add 0.5s of silence if audio is shorter than the silence_padding window.
+
+    Whisper does not require this strictly, but aligning behavior with Parakeet
+    avoids edge-case crashes on extremely short inputs and makes comparisons easier.
+    """
+    import numpy as np
+
+    audio_duration = len(audio_array) / sample_rate
+    if audio_duration < VAD_CONFIG["silence_padding"]:
+        silence_samples = int(sample_rate * VAD_CONFIG["silence_padding"])
+        silence = np.zeros(silence_samples, dtype=np.float32)
+        return np.concatenate([audio_array, silence])
+    return audio_array
+
+
 @app.cls(
     gpu="A10G",
     timeout=5 * MINUTES,
     scaledown_window=5 * MINUTES,
-    allow_concurrent_inputs=6,
     image=image,
-    volumes={MODELS_DIR: volume},
+    volumes={CACHE_PATH: model_cache, UPLOADS_PATH: upload_volume},
 )
-class Transcriber:
+@modal.concurrent(max_inputs=10)
+class TranscriberWhisperLive:
+    """Live transcriber class for small audio segments (A10G).
+
+    Mirrors the Parakeet live class API but uses Faster-Whisper under the hood.
+    """
+
     @modal.enter()
     def enter(self):
         import faster_whisper
@@ -71,23 +187,200 @@ class Transcriber:
             device=self.device,
             compute_type=MODEL_COMPUTE_TYPE,
             num_workers=MODEL_NUM_WORKERS,
-            download_root=MODELS_DIR,
+            download_root=CACHE_PATH,
             local_files_only=True,
         )
+        print(f"Model is on device: {self.device}")
 
     @modal.method()
     def transcribe_segment(
         self,
-        audio_data: str,
-        audio_suffix: str,
-        language: str,
+        filename: str,
+        language: str = "en",
     ):
-        with tempfile.NamedTemporaryFile("wb+", suffix=f".{audio_suffix}") as fp:
-            fp.write(audio_data)
+        """Transcribe a single uploaded audio file by filename."""
+        upload_volume.reload()
+
+        file_path = f"{UPLOADS_PATH}/{filename}"
+        if not os.path.exists(file_path):
+            raise FileNotFoundError(f"File not found: {file_path}")
+
+        with self.lock:
+            with NoStdStreams():
+                segments, _ = self.model.transcribe(
+                    file_path,
+                    language=language,
+                    beam_size=5,
+                    word_timestamps=True,
+                    vad_filter=True,
+                    vad_parameters={"min_silence_duration_ms": 500},
+                )
+
+        segments = list(segments)
+        text = "".join(segment.text for segment in segments).strip()
+        words = [
+            {
+                "word": word.word,
+                "start": round(float(word.start), 2),
+                "end": round(float(word.end), 2),
+            }
+            for segment in segments
+            for word in segment.words
+        ]
+
+        return {"text": text, "words": words}
+
+    @modal.method()
+    def transcribe_batch(
+        self,
+        filenames: list[str],
+        language: str = "en",
+    ):
+        """Transcribe multiple uploaded audio files and return per-file results."""
+        upload_volume.reload()
+
+        results = []
+        for filename in filenames:
+            file_path = f"{UPLOADS_PATH}/{filename}"
+            if not os.path.exists(file_path):
+                raise FileNotFoundError(f"Batch file not found: {file_path}")
+
+            with self.lock:
+                with NoStdStreams():
+                    segments, _ = self.model.transcribe(
+                        file_path,
+                        language=language,
+                        beam_size=5,
+                        word_timestamps=True,
+                        vad_filter=True,
+                        vad_parameters={"min_silence_duration_ms": 500},
+                    )
+
+            segments = list(segments)
+            text = "".join(seg.text for seg in segments).strip()
+            words = [
+                {
+                    "word": w.word,
+                    "start": round(float(w.start), 2),
+                    "end": round(float(w.end), 2),
+                }
+                for seg in segments
+                for w in seg.words
+            ]
+
+            results.append(
+                {
+                    "filename": filename,
+                    "text": text,
+                    "words": words,
+                }
+            )
+
+        return results
+
+
+@app.cls(
+    gpu="L40S",
+    timeout=15 * MINUTES,
+    image=image,
+    volumes={CACHE_PATH: model_cache, UPLOADS_PATH: upload_volume},
+)
+class TranscriberWhisperFile:
+    """File transcriber for larger/longer audio, using VAD-driven batching (L40S)."""
+
+    @modal.enter()
+    def enter(self):
+        import faster_whisper
+        import torch
+        from silero_vad import load_silero_vad
+
+        self.lock = threading.Lock()
+        self.use_gpu = torch.cuda.is_available()
+        self.device = "cuda" if self.use_gpu else "cpu"
+        self.model = faster_whisper.WhisperModel(
+            MODEL_NAME,
+            device=self.device,
+            compute_type=MODEL_COMPUTE_TYPE,
+            num_workers=MODEL_NUM_WORKERS,
+            download_root=CACHE_PATH,
+            local_files_only=True,
+        )
+        self.vad_model = load_silero_vad(onnx=False)
+
+    @modal.method()
+    def transcribe_segment(
+        self, filename: str, timestamp_offset: float = 0.0, language: str = "en"
+    ):
+        import librosa
+        import numpy as np
+        from silero_vad import VADIterator
+
+        def vad_segments(
+            audio_array,
+            sample_rate: int = SAMPLERATE,
+            window_size: int = VAD_CONFIG["window_size"],
+        ) -> Generator[TimeSegment, None, None]:
+            """Generate speech segments as TimeSegment using Silero VAD."""
+            iterator = VADIterator(self.vad_model, sampling_rate=sample_rate)
+            start = None
+            for i in range(0, len(audio_array), window_size):
+                chunk = audio_array[i : i + window_size]
+                if len(chunk) < window_size:
+                    chunk = np.pad(
+                        chunk, (0, window_size - len(chunk)), mode="constant"
+                    )
+                speech = iterator(chunk)
+                if not speech:
+                    continue
+                if "start" in speech:
+                    start = speech["start"]
+                    continue
+                if "end" in speech and start is not None:
+                    end = speech["end"]
+                    yield TimeSegment(
+                        start / float(SAMPLERATE), end / float(SAMPLERATE)
+                    )
+                    start = None
+            iterator.reset_states()
+
+        upload_volume.reload()
+        file_path = f"{UPLOADS_PATH}/{filename}"
+        if not os.path.exists(file_path):
+            raise FileNotFoundError(f"File not found: {file_path}")
+
+        audio_array, _sr = librosa.load(file_path, sr=SAMPLERATE, mono=True)
+
+        # Batch segments up to ~30s windows by merging contiguous VAD segments
+        merged_batches: list[TimeSegment] = []
+        batch_start = None
+        batch_end = None
+        max_duration = VAD_CONFIG["batch_max_duration"]
+        for segment in vad_segments(audio_array):
+            seg_start, seg_end = segment.start, segment.end
+            if batch_start is None:
+                batch_start, batch_end = seg_start, seg_end
+                continue
+            if seg_end - batch_start <= max_duration:
+                batch_end = seg_end
+            else:
+                merged_batches.append(TimeSegment(batch_start, batch_end))
+                batch_start, batch_end = seg_start, seg_end
+        if batch_start is not None and batch_end is not None:
+            merged_batches.append(TimeSegment(batch_start, batch_end))
+
+        all_text = []
+        all_words = []
+
+        for segment in merged_batches:
+            start_time, end_time = segment.start, segment.end
+            s_idx = int(start_time * SAMPLERATE)
+            e_idx = int(end_time * SAMPLERATE)
+            segment = audio_array[s_idx:e_idx]
+            segment = pad_audio(segment, SAMPLERATE)
 
             with self.lock:
                 segments, _ = self.model.transcribe(
-                    fp.name,
+                    segment,
                     language=language,
                     beam_size=5,
                     word_timestamps=True,
@@ -96,66 +389,220 @@ class Transcriber:
                 )
 
             segments = list(segments)
-            text = "".join(segment.text for segment in segments)
+            text = "".join(seg.text for seg in segments).strip()
             words = [
-                {"word": word.word, "start": word.start, "end": word.end}
-                for segment in segments
-                for word in segment.words
+                {
+                    "word": w.word,
+                    "start": round(float(w.start) + start_time + timestamp_offset, 2),
+                    "end": round(float(w.end) + start_time + timestamp_offset, 2),
+                }
+                for seg in segments
+                for w in seg.words
             ]
+            if text:
+                all_text.append(text)
+            all_words.extend(words)
 
-            return {"text": text, "words": words}
+        return {"text": " ".join(all_text), "words": all_words}
+
+
+def detect_audio_format(url: str, headers: dict) -> str:
+    from urllib.parse import urlparse
+
+    from fastapi import HTTPException
+
+    url_path = urlparse(url).path
+    for ext in SUPPORTED_FILE_EXTENSIONS:
+        if url_path.lower().endswith(f".{ext}"):
+            return ext
+
+    content_type = headers.get("content-type", "").lower()
+    if "audio/mpeg" in content_type or "audio/mp3" in content_type:
+        return "mp3"
+    if "audio/wav" in content_type:
+        return "wav"
+    if "audio/mp4" in content_type:
+        return "mp4"
+
+    raise HTTPException(
+        status_code=400,
+        detail=(
+            f"Unsupported audio format for URL. Supported extensions: {', '.join(SUPPORTED_FILE_EXTENSIONS)}"
+        ),
+    )
+
+
+def download_audio_to_volume(audio_file_url: str) -> tuple[str, str]:
+    import requests
+    from fastapi import HTTPException
+
+    response = requests.head(audio_file_url, allow_redirects=True)
+    if response.status_code == 404:
+        raise HTTPException(status_code=404, detail="Audio file not found")
+
+    response = requests.get(audio_file_url, allow_redirects=True)
+    response.raise_for_status()
+
+    audio_suffix = detect_audio_format(audio_file_url, response.headers)
+    unique_filename = f"{uuid.uuid4()}.{audio_suffix}"
+    file_path = f"{UPLOADS_PATH}/{unique_filename}"
+
+    with open(file_path, "wb") as f:
+        f.write(response.content)
+
+    upload_volume.commit()
+    return unique_filename, audio_suffix
 
 
 @app.function(
     scaledown_window=60,
-    timeout=60,
-    allow_concurrent_inputs=40,
+    timeout=600,
     secrets=[
         modal.Secret.from_name("reflector-gpu"),
     ],
-    volumes={MODELS_DIR: volume},
+    volumes={CACHE_PATH: model_cache, UPLOADS_PATH: upload_volume},
+    image=image,
 )
+@modal.concurrent(max_inputs=40)
 @modal.asgi_app()
 def web():
-    from fastapi import Body, Depends, FastAPI, HTTPException, UploadFile, status
+    from fastapi import (
+        Body,
+        Depends,
+        FastAPI,
+        Form,
+        HTTPException,
+        UploadFile,
+        status,
+    )
     from fastapi.security import OAuth2PasswordBearer
-    from typing_extensions import Annotated
 
-    transcriber = Transcriber()
+    transcriber_live = TranscriberWhisperLive()
+    transcriber_file = TranscriberWhisperFile()
 
     app = FastAPI()
 
     oauth2_scheme = OAuth2PasswordBearer(tokenUrl="token")
 
-    supported_file_types = ["mp3", "mp4", "mpeg", "mpga", "m4a", "wav", "webm"]
-
     def apikey_auth(apikey: str = Depends(oauth2_scheme)):
-        if apikey != os.environ["REFLECTOR_GPU_APIKEY"]:
-            raise HTTPException(
-                status_code=status.HTTP_401_UNAUTHORIZED,
-                detail="Invalid API key",
-                headers={"WWW-Authenticate": "Bearer"},
-            )
+        if apikey == os.environ["REFLECTOR_GPU_APIKEY"]:
+            return
+        raise HTTPException(
+            status_code=status.HTTP_401_UNAUTHORIZED,
+            detail="Invalid API key",
+            headers={"WWW-Authenticate": "Bearer"},
+        )
 
-    class TranscriptResponse(BaseModel):
-        result: dict
+    class TranscriptResponse(dict):
+        pass
 
     @app.post("/v1/audio/transcriptions", dependencies=[Depends(apikey_auth)])
     def transcribe(
-        file: UploadFile,
-        model: str = "whisper-1",
-        language: Annotated[str, Body(...)] = "en",
-    ) -> TranscriptResponse:
-        audio_data = file.file.read()
-        audio_suffix = file.filename.split(".")[-1]
-        assert audio_suffix in supported_file_types
+        file: UploadFile = None,
+        files: list[UploadFile] | None = None,
+        model: str = Form(MODEL_NAME),
+        language: str = Form("en"),
+        batch: bool = Form(False),
+    ):
+        if not file and not files:
+            raise HTTPException(
+                status_code=400, detail="Either 'file' or 'files' parameter is required"
+            )
+        if batch and not files:
+            raise HTTPException(
+                status_code=400, detail="Batch transcription requires 'files'"
+            )
 
-        func = transcriber.transcribe_segment.spawn(
-            audio_data=audio_data,
-            audio_suffix=audio_suffix,
-            language=language,
-        )
-        result = func.get()
-        return result
+        upload_files = [file] if file else files
+
+        uploaded_filenames: list[str] = []
+        for upload_file in upload_files:
+            audio_suffix = upload_file.filename.split(".")[-1]
+            if audio_suffix not in SUPPORTED_FILE_EXTENSIONS:
+                raise HTTPException(
+                    status_code=400,
+                    detail=(
+                        f"Unsupported audio format. Supported extensions: {', '.join(SUPPORTED_FILE_EXTENSIONS)}"
+                    ),
+                )
+
+            unique_filename = f"{uuid.uuid4()}.{audio_suffix}"
+            file_path = f"{UPLOADS_PATH}/{unique_filename}"
+            with open(file_path, "wb") as f:
+                content = upload_file.file.read()
+                f.write(content)
+            uploaded_filenames.append(unique_filename)
+
+        upload_volume.commit()
+
+        try:
+            if batch and len(upload_files) > 1:
+                func = transcriber_live.transcribe_batch.spawn(
+                    filenames=uploaded_filenames,
+                    language=language,
+                )
+                results = func.get()
+                return {"results": results}
+
+            results = []
+            for filename in uploaded_filenames:
+                func = transcriber_live.transcribe_segment.spawn(
+                    filename=filename,
+                    language=language,
+                )
+                result = func.get()
+                result["filename"] = filename
+                results.append(result)
+
+            return {"results": results} if len(results) > 1 else results[0]
+        finally:
+            for filename in uploaded_filenames:
+                try:
+                    file_path = f"{UPLOADS_PATH}/{filename}"
+                    os.remove(file_path)
+                except Exception:
+                    pass
+            upload_volume.commit()
+
+    @app.post("/v1/audio/transcriptions-from-url", dependencies=[Depends(apikey_auth)])
+    def transcribe_from_url(
+        audio_file_url: str = Body(
+            ..., description="URL of the audio file to transcribe"
+        ),
+        model: str = Body(MODEL_NAME),
+        language: str = Body("en"),
+        timestamp_offset: float = Body(0.0),
+    ):
+        unique_filename, _audio_suffix = download_audio_to_volume(audio_file_url)
+        try:
+            func = transcriber_file.transcribe_segment.spawn(
+                filename=unique_filename,
+                timestamp_offset=timestamp_offset,
+                language=language,
+            )
+            result = func.get()
+            return result
+        finally:
+            try:
+                file_path = f"{UPLOADS_PATH}/{unique_filename}"
+                os.remove(file_path)
+                upload_volume.commit()
+            except Exception:
+                pass
 
     return app
+
+
+class NoStdStreams:
+    def __init__(self):
+        self.devnull = open(os.devnull, "w")
+
+    def __enter__(self):
+        self._stdout, self._stderr = sys.stdout, sys.stderr
+        self._stdout.flush()
+        self._stderr.flush()
+        sys.stdout, sys.stderr = self.devnull, self.devnull
+
+    def __exit__(self, exc_type, exc_value, traceback):
+        sys.stdout, sys.stderr = self._stdout, self._stderr
+        self.devnull.close()

server/gpu/modal_deployments/reflector_transcriber_parakeet.py

@@ -3,7 +3,7 @@ import os
 import sys
 import threading
 import uuid
-from typing import Mapping, NewType
+from typing import Generator, Mapping, NamedTuple, NewType, TypedDict
 from urllib.parse import urlparse
 
 import modal
@@ -14,10 +14,7 @@ SAMPLERATE = 16000
 UPLOADS_PATH = "/uploads"
 CACHE_PATH = "/cache"
 VAD_CONFIG = {
-    "max_segment_duration": 30.0,
-    "batch_max_files": 10,
-    "batch_max_duration": 5.0,
-    "min_segment_duration": 0.02,
+    "batch_max_duration": 30.0,
     "silence_padding": 0.5,
     "window_size": 512,
 }
@@ -25,6 +22,37 @@ VAD_CONFIG = {
 ParakeetUniqFilename = NewType("ParakeetUniqFilename", str)
 AudioFileExtension = NewType("AudioFileExtension", str)
 
+
+class TimeSegment(NamedTuple):
+    """Represents a time segment with start and end times."""
+
+    start: float
+    end: float
+
+
+class AudioSegment(NamedTuple):
+    """Represents an audio segment with timing and audio data."""
+
+    start: float
+    end: float
+    audio: any
+
+
+class TranscriptResult(NamedTuple):
+    """Represents a transcription result with text and word timings."""
+
+    text: str
+    words: list["WordTiming"]
+
+
+class WordTiming(TypedDict):
+    """Represents a word with its timing information."""
+
+    word: str
+    start: float
+    end: float
+
+
 app = modal.App("reflector-transcriber-parakeet")
 
 # Volume for caching model weights
@@ -170,12 +198,14 @@ class TranscriberParakeetLive:
                 (output,) = self.model.transcribe([padded_audio], timestamps=True)
 
         text = output.text.strip()
-        words = [
-            {
-                "word": word_info["word"],
-                "start": round(word_info["start"], 2),
-                "end": round(word_info["end"], 2),
-            }
+        words: list[WordTiming] = [
+            WordTiming(
+                # XXX the space added here is to match the output of whisper
+                # whisper add space to each words, while parakeet don't
+                word=word_info["word"] + " ",
+                start=round(word_info["start"], 2),
+                end=round(word_info["end"], 2),
+            )
             for word_info in output.timestamp["word"]
         ]
 
@@ -211,12 +241,12 @@ class TranscriberParakeetLive:
         for i, (filename, output) in enumerate(zip(filenames, outputs)):
             text = output.text.strip()
 
-            words = [
-                {
-                    "word": word_info["word"],
-                    "start": round(word_info["start"], 2),
-                    "end": round(word_info["end"], 2),
-                }
+            words: list[WordTiming] = [
+                WordTiming(
+                    word=word_info["word"] + " ",
+                    start=round(word_info["start"], 2),
+                    end=round(word_info["end"], 2),
+                )
                 for word_info in output.timestamp["word"]
             ]
 
@@ -271,7 +301,9 @@ class TranscriberParakeetFile:
             audio_array, sample_rate = librosa.load(file_path, sr=SAMPLERATE, mono=True)
             return audio_array
 
-        def vad_segment_generator(audio_array):
+        def vad_segment_generator(
+            audio_array,
+        ) -> Generator[TimeSegment, None, None]:
             """Generate speech segments using VAD with start/end sample indices"""
             vad_iterator = VADIterator(self.vad_model, sampling_rate=SAMPLERATE)
             window_size = VAD_CONFIG["window_size"]
@@ -297,107 +329,121 @@ class TranscriberParakeetFile:
                     start_time = start / float(SAMPLERATE)
                     end_time = end / float(SAMPLERATE)
 
-                    # Extract the actual audio segment
-                    audio_segment = audio_array[start:end]
-
-                    yield (start_time, end_time, audio_segment)
+                    yield TimeSegment(start_time, end_time)
                     start = None
 
             vad_iterator.reset_states()
 
-        def vad_segment_filter(segments):
-            """Filter VAD segments by duration and chunk large segments"""
-            min_dur = VAD_CONFIG["min_segment_duration"]
-            max_dur = VAD_CONFIG["max_segment_duration"]
+        def batch_speech_segments(
+            segments: Generator[TimeSegment, None, None], max_duration: int
+        ) -> Generator[TimeSegment, None, None]:
+            """
+            Input segments:
+              [0-2] [3-5] [6-8] [10-11] [12-15] [17-19] [20-22]
 
-            for start_time, end_time, audio_segment in segments:
-                segment_duration = end_time - start_time
+                                  ↓ (max_duration=10)
 
-                # Skip very small segments
-                if segment_duration < min_dur:
+              Output batches:
+              [0-8]           [10-19]          [20-22]
+
+            Note: silences are kept for better transcription, previous implementation was
+            passing segments separatly, but the output was less accurate.
+            """
+            batch_start_time = None
+            batch_end_time = None
+
+            for segment in segments:
+                start_time, end_time = segment.start, segment.end
+                if batch_start_time is None or batch_end_time is None:
+                    batch_start_time = start_time
+                    batch_end_time = end_time
                     continue
 
-                # If segment is within max duration, yield as-is
-                if segment_duration <= max_dur:
-                    yield (start_time, end_time, audio_segment)
+                total_duration = end_time - batch_start_time
+
+                if total_duration <= max_duration:
+                    batch_end_time = end_time
                     continue
 
-                # Chunk large segments into smaller pieces
-                chunk_samples = int(max_dur * SAMPLERATE)
-                current_start = start_time
+                yield TimeSegment(batch_start_time, batch_end_time)
+                batch_start_time = start_time
+                batch_end_time = end_time
 
-                for chunk_offset in range(0, len(audio_segment), chunk_samples):
-                    chunk_audio = audio_segment[
-                        chunk_offset : chunk_offset + chunk_samples
-                    ]
-                    if len(chunk_audio) == 0:
-                        break
+            if batch_start_time is None or batch_end_time is None:
+                return
 
-                    chunk_duration = len(chunk_audio) / float(SAMPLERATE)
-                    chunk_end = current_start + chunk_duration
+            yield TimeSegment(batch_start_time, batch_end_time)
 
-                    # Only yield chunks that meet minimum duration
-                    if chunk_duration >= min_dur:
-                        yield (current_start, chunk_end, chunk_audio)
+        def batch_segment_to_audio_segment(
+            segments: Generator[TimeSegment, None, None],
+            audio_array,
+        ) -> Generator[AudioSegment, None, None]:
+            """Extract audio segments and apply padding for Parakeet compatibility.
 
-                    current_start = chunk_end
+            Uses pad_audio to ensure segments are at least 0.5s long, preventing
+            Parakeet crashes. This padding may cause slight timing overlaps between
+            segments, which are corrected by enforce_word_timing_constraints.
+            """
+            for segment in segments:
+                start_time, end_time = segment.start, segment.end
+                start_sample = int(start_time * SAMPLERATE)
+                end_sample = int(end_time * SAMPLERATE)
+                audio_segment = audio_array[start_sample:end_sample]
 
-        def batch_segments(segments, max_files=10, max_duration=5.0):
-            batch = []
-            batch_duration = 0.0
+                padded_segment = pad_audio(audio_segment, SAMPLERATE)
 
-            for start_time, end_time, audio_segment in segments:
-                segment_duration = end_time - start_time
+                yield AudioSegment(start_time, end_time, padded_segment)
 
-                if segment_duration < VAD_CONFIG["silence_padding"]:
-                    silence_samples = int(
-                        (VAD_CONFIG["silence_padding"] - segment_duration) * SAMPLERATE
-                    )
-                    padding = np.zeros(silence_samples, dtype=np.float32)
-                    audio_segment = np.concatenate([audio_segment, padding])
-                    segment_duration = VAD_CONFIG["silence_padding"]
-
-                batch.append((start_time, end_time, audio_segment))
-                batch_duration += segment_duration
-
-                if len(batch) >= max_files or batch_duration >= max_duration:
-                    yield batch
-                    batch = []
-                    batch_duration = 0.0
-
-            if batch:
-                yield batch
-
-        def transcribe_batch(model, audio_segments):
+        def transcribe_batch(model, audio_segments: list) -> list:
             with NoStdStreams():
                 outputs = model.transcribe(audio_segments, timestamps=True)
             return outputs
 
+        def enforce_word_timing_constraints(
+            words: list[WordTiming],
+        ) -> list[WordTiming]:
+            """Enforce that word end times don't exceed the start time of the next word.
+
+            Due to silence padding added in batch_segment_to_audio_segment for better
+            transcription accuracy, word timings from different segments may overlap.
+            This function ensures there are no overlaps by adjusting end times.
+            """
+            if len(words) <= 1:
+                return words
+
+            enforced_words = []
+            for i, word in enumerate(words):
+                enforced_word = word.copy()
+
+                if i < len(words) - 1:
+                    next_start = words[i + 1]["start"]
+                    if enforced_word["end"] > next_start:
+                        enforced_word["end"] = next_start
+
+                enforced_words.append(enforced_word)
+
+            return enforced_words
+
         def emit_results(
-            results,
-            segments_info,
-            batch_index,
-            total_batches,
-        ):
+            results: list,
+            segments_info: list[AudioSegment],
+        ) -> Generator[TranscriptResult, None, None]:
             """Yield transcribed text and word timings from model output, adjusting timestamps to absolute positions."""
-            for i, (output, (start_time, end_time, _)) in enumerate(
-                zip(results, segments_info)
-            ):
+            for i, (output, segment) in enumerate(zip(results, segments_info)):
+                start_time, end_time = segment.start, segment.end
                 text = output.text.strip()
-                words = [
-                    {
-                        "word": word_info["word"],
-                        "start": round(
+                words: list[WordTiming] = [
+                    WordTiming(
+                        word=word_info["word"] + " ",
+                        start=round(
                             word_info["start"] + start_time + timestamp_offset, 2
                         ),
-                        "end": round(
-                            word_info["end"] + start_time + timestamp_offset, 2
-                        ),
-                    }
+                        end=round(word_info["end"] + start_time + timestamp_offset, 2),
+                    )
                     for word_info in output.timestamp["word"]
                 ]
 
-                yield text, words
+                yield TranscriptResult(text, words)
 
         upload_volume.reload()
 
@@ -407,41 +453,31 @@ class TranscriberParakeetFile:
 
         audio_array = load_and_convert_audio(file_path)
         total_duration = len(audio_array) / float(SAMPLERATE)
-        processed_duration = 0.0
 
-        all_text_parts = []
-        all_words = []
+        all_text_parts: list[str] = []
+        all_words: list[WordTiming] = []
 
         raw_segments = vad_segment_generator(audio_array)
-        filtered_segments = vad_segment_filter(raw_segments)
-        batches = batch_segments(
-            filtered_segments,
-            VAD_CONFIG["batch_max_files"],
+        speech_segments = batch_speech_segments(
+            raw_segments,
             VAD_CONFIG["batch_max_duration"],
         )
+        audio_segments = batch_segment_to_audio_segment(speech_segments, audio_array)
 
-        batch_index = 0
-        total_batches = max(
-            1, int(total_duration / VAD_CONFIG["batch_max_duration"]) + 1
-        )
+        for batch in audio_segments:
+            audio_segment = batch.audio
+            results = transcribe_batch(self.model, [audio_segment])
 
-        for batch in batches:
-            batch_index += 1
-            audio_segments = [seg[2] for seg in batch]
-            results = transcribe_batch(self.model, audio_segments)
-
-            for text, words in emit_results(
+            for result in emit_results(
                 results,
-                batch,
-                batch_index,
-                total_batches,
+                [batch],
             ):
-                if not text:
+                if not result.text:
                     continue
-                all_text_parts.append(text)
-                all_words.extend(words)
+                all_text_parts.append(result.text)
+                all_words.extend(result.words)
 
-            processed_duration += sum(len(seg[2]) / float(SAMPLERATE) for seg in batch)
+        all_words = enforce_word_timing_constraints(all_words)
 
         combined_text = " ".join(all_text_parts)
         return {"text": combined_text, "words": all_words}

server/migrations/versions/0194f65cd6d3_add_webhook_fields_to_rooms.py

@@ -0,0 +1,36 @@
+"""Add webhook fields to rooms
+
+Revision ID: 0194f65cd6d3
+Revises: 5a8907fd1d78
+Create Date: 2025-08-27 09:03:19.610995
+
+"""
+
+from typing import Sequence, Union
+
+import sqlalchemy as sa
+from alembic import op
+
+# revision identifiers, used by Alembic.
+revision: str = "0194f65cd6d3"
+down_revision: Union[str, None] = "5a8907fd1d78"
+branch_labels: Union[str, Sequence[str], None] = None
+depends_on: Union[str, Sequence[str], None] = None
+
+
+def upgrade() -> None:
+    # ### commands auto generated by Alembic - please adjust! ###
+    with op.batch_alter_table("room", schema=None) as batch_op:
+        batch_op.add_column(sa.Column("webhook_url", sa.String(), nullable=True))
+        batch_op.add_column(sa.Column("webhook_secret", sa.String(), nullable=True))
+
+    # ### end Alembic commands ###
+
+
+def downgrade() -> None:
+    # ### commands auto generated by Alembic - please adjust! ###
+    with op.batch_alter_table("room", schema=None) as batch_op:
+        batch_op.drop_column("webhook_secret")
+        batch_op.drop_column("webhook_url")
+
+    # ### end Alembic commands ###

server/migrations/versions/5a8907fd1d78_add_cascade_delete_to_meeting_consent_.py

@@ -0,0 +1,50 @@
+"""add cascade delete to meeting consent foreign key
+
+Revision ID: 5a8907fd1d78
+Revises: 0ab2d7ffaa16
+Create Date: 2025-08-26 17:26:50.945491
+
+"""
+
+from typing import Sequence, Union
+
+from alembic import op
+
+# revision identifiers, used by Alembic.
+revision: str = "5a8907fd1d78"
+down_revision: Union[str, None] = "0ab2d7ffaa16"
+branch_labels: Union[str, Sequence[str], None] = None
+depends_on: Union[str, Sequence[str], None] = None
+
+
+def upgrade() -> None:
+    # ### commands auto generated by Alembic - please adjust! ###
+    with op.batch_alter_table("meeting_consent", schema=None) as batch_op:
+        batch_op.drop_constraint(
+            batch_op.f("meeting_consent_meeting_id_fkey"), type_="foreignkey"
+        )
+        batch_op.create_foreign_key(
+            batch_op.f("meeting_consent_meeting_id_fkey"),
+            "meeting",
+            ["meeting_id"],
+            ["id"],
+            ondelete="CASCADE",
+        )
+
+    # ### end Alembic commands ###
+
+
+def downgrade() -> None:
+    # ### commands auto generated by Alembic - please adjust! ###
+    with op.batch_alter_table("meeting_consent", schema=None) as batch_op:
+        batch_op.drop_constraint(
+            batch_op.f("meeting_consent_meeting_id_fkey"), type_="foreignkey"
+        )
+        batch_op.create_foreign_key(
+            batch_op.f("meeting_consent_meeting_id_fkey"),
+            "meeting",
+            ["meeting_id"],
+            ["id"],
+        )
+
+    # ### end Alembic commands ###

server/migrations/versions/61882a919591_webhook_url_and_secret_null_by_default.py

@@ -0,0 +1,28 @@
+"""webhook url and secret null by default
+
+
+Revision ID: 61882a919591
+Revises: 0194f65cd6d3
+Create Date: 2025-08-29 11:46:36.738091
+
+"""
+
+from typing import Sequence, Union
+
+# revision identifiers, used by Alembic.
+revision: str = "61882a919591"
+down_revision: Union[str, None] = "0194f65cd6d3"
+branch_labels: Union[str, Sequence[str], None] = None
+depends_on: Union[str, Sequence[str], None] = None
+
+
+def upgrade() -> None:
+    # ### commands auto generated by Alembic - please adjust! ###
+    pass
+    # ### end Alembic commands ###
+
+
+def downgrade() -> None:
+    # ### commands auto generated by Alembic - please adjust! ###
+    pass
+    # ### end Alembic commands ###

server/reflector/asynctask.py

@@ -0,0 +1,27 @@
+import asyncio
+import functools
+
+from reflector.db import get_database
+
+
+def asynctask(f):
+    @functools.wraps(f)
+    def wrapper(*args, **kwargs):
+        async def run_with_db():
+            database = get_database()
+            await database.connect()
+            try:
+                return await f(*args, **kwargs)
+            finally:
+                await database.disconnect()
+
+        coro = run_with_db()
+        try:
+            loop = asyncio.get_running_loop()
+        except RuntimeError:
+            loop = None
+        if loop and loop.is_running():
+            return loop.run_until_complete(coro)
+        return asyncio.run(coro)
+
+    return wrapper

server/reflector/db/meetings.py

@@ -54,7 +54,12 @@ meeting_consent = sa.Table(
     "meeting_consent",
     metadata,
     sa.Column("id", sa.String, primary_key=True),
-    sa.Column("meeting_id", sa.String, sa.ForeignKey("meeting.id"), nullable=False),
+    sa.Column(
+        "meeting_id",
+        sa.String,
+        sa.ForeignKey("meeting.id", ondelete="CASCADE"),
+        nullable=False,
+    ),
     sa.Column("user_id", sa.String),
     sa.Column("consent_given", sa.Boolean, nullable=False),
     sa.Column("consent_timestamp", sa.DateTime(timezone=True), nullable=False),

server/reflector/db/rooms.py

@@ -1,3 +1,4 @@
+import secrets
 from datetime import datetime, timezone
 from sqlite3 import IntegrityError
 from typing import Literal
@@ -40,6 +41,8 @@ rooms = sqlalchemy.Table(
     sqlalchemy.Column(
         "is_shared", sqlalchemy.Boolean, nullable=False, server_default=false()
     ),
+    sqlalchemy.Column("webhook_url", sqlalchemy.String, nullable=True),
+    sqlalchemy.Column("webhook_secret", sqlalchemy.String, nullable=True),
     sqlalchemy.Index("idx_room_is_shared", "is_shared"),
 )
 
@@ -59,6 +62,8 @@ class Room(BaseModel):
         "none", "prompt", "automatic", "automatic-2nd-participant"
     ] = "automatic-2nd-participant"
     is_shared: bool = False
+    webhook_url: str | None = None
+    webhook_secret: str | None = None
 
 
 class RoomController:
@@ -107,10 +112,15 @@ class RoomController:
         recording_type: str,
         recording_trigger: str,
         is_shared: bool,
+        webhook_url: str = "",
+        webhook_secret: str = "",
     ):
         """
         Add a new room
         """
+        if webhook_url and not webhook_secret:
+            webhook_secret = secrets.token_urlsafe(32)
+
         room = Room(
             name=name,
             user_id=user_id,
@@ -122,6 +132,8 @@ class RoomController:
             recording_type=recording_type,
             recording_trigger=recording_trigger,
             is_shared=is_shared,
+            webhook_url=webhook_url,
+            webhook_secret=webhook_secret,
         )
         query = rooms.insert().values(**room.model_dump())
         try:
@@ -134,6 +146,9 @@ class RoomController:
         """
         Update a room fields with key/values in values
         """
+        if values.get("webhook_url") and not values.get("webhook_secret"):
+            values["webhook_secret"] = secrets.token_urlsafe(32)
+
         query = rooms.update().where(rooms.c.id == room.id).values(**values)
         try:
             await get_database().execute(query)

server/reflector/db/search.py

@@ -8,12 +8,14 @@ from typing import Annotated, Any, Dict, Iterator
 
 import sqlalchemy
 import webvtt
+from databases.interfaces import Record as DbRecord
 from fastapi import HTTPException
 from pydantic import (
     BaseModel,
     Field,
     NonNegativeFloat,
     NonNegativeInt,
+    TypeAdapter,
     ValidationError,
     constr,
     field_serializer,
@@ -24,6 +26,7 @@ from reflector.db.rooms import rooms
 from reflector.db.transcripts import SourceKind, transcripts
 from reflector.db.utils import is_postgresql
 from reflector.logger import logger
+from reflector.utils.string import NonEmptyString, try_parse_non_empty_string
 
 DEFAULT_SEARCH_LIMIT = 20
 SNIPPET_CONTEXT_LENGTH = 50  # Characters before/after match to include
@@ -31,12 +34,13 @@ DEFAULT_SNIPPET_MAX_LENGTH = NonNegativeInt(150)
 DEFAULT_MAX_SNIPPETS = NonNegativeInt(3)
 LONG_SUMMARY_MAX_SNIPPETS = 2
 
-SearchQueryBase = constr(min_length=0, strip_whitespace=True)
+SearchQueryBase = constr(min_length=1, strip_whitespace=True)
 SearchLimitBase = Annotated[int, Field(ge=1, le=100)]
 SearchOffsetBase = Annotated[int, Field(ge=0)]
 SearchTotalBase = Annotated[int, Field(ge=0)]
 
 SearchQuery = Annotated[SearchQueryBase, Field(description="Search query text")]
+search_query_adapter = TypeAdapter(SearchQuery)
 SearchLimit = Annotated[SearchLimitBase, Field(description="Results per page")]
 SearchOffset = Annotated[
     SearchOffsetBase, Field(description="Number of results to skip")
@@ -88,7 +92,7 @@ class WebVTTProcessor:
     @staticmethod
     def generate_snippets(
         webvtt_content: WebVTTContent,
-        query: str,
+        query: SearchQuery,
         max_snippets: NonNegativeInt = DEFAULT_MAX_SNIPPETS,
     ) -> list[str]:
         """Generate snippets from WebVTT content."""
@@ -125,7 +129,7 @@ class SnippetCandidate:
 class SearchParameters(BaseModel):
     """Validated search parameters for full-text search."""
 
-    query_text: SearchQuery
+    query_text: SearchQuery | None = None
     limit: SearchLimit = DEFAULT_SEARCH_LIMIT
     offset: SearchOffset = 0
     user_id: str | None = None
@@ -199,15 +203,13 @@ class SnippetGenerator:
             prev_start = start
 
     @staticmethod
-    def count_matches(text: str, query: str) -> NonNegativeInt:
+    def count_matches(text: str, query: SearchQuery) -> NonNegativeInt:
         """Count total number of matches for a query in text."""
         ZERO = NonNegativeInt(0)
         if not text:
             logger.warning("Empty text for search query in count_matches")
             return ZERO
-        if not query:
-            logger.warning("Empty query for search text in count_matches")
-            return ZERO
+        assert query is not None
         return NonNegativeInt(
             sum(1 for _ in SnippetGenerator.find_all_matches(text, query))
         )
@@ -243,13 +245,14 @@ class SnippetGenerator:
     @staticmethod
     def generate(
         text: str,
-        query: str,
+        query: SearchQuery,
         max_length: NonNegativeInt = DEFAULT_SNIPPET_MAX_LENGTH,
         max_snippets: NonNegativeInt = DEFAULT_MAX_SNIPPETS,
     ) -> list[str]:
         """Generate snippets from text."""
-        if not text or not query:
-            logger.warning("Empty text or query for generate_snippets")
+        assert query is not None
+        if not text:
+            logger.warning("Empty text for generate_snippets")
             return []
 
         candidates = (
@@ -270,7 +273,7 @@ class SnippetGenerator:
     @staticmethod
     def from_summary(
         summary: str,
-        query: str,
+        query: SearchQuery,
         max_snippets: NonNegativeInt = LONG_SUMMARY_MAX_SNIPPETS,
     ) -> list[str]:
         """Generate snippets from summary text."""
@@ -278,9 +281,9 @@ class SnippetGenerator:
 
     @staticmethod
     def combine_sources(
-        summary: str | None,
+        summary: NonEmptyString | None,
         webvtt: WebVTTContent | None,
-        query: str,
+        query: SearchQuery,
         max_total: NonNegativeInt = DEFAULT_MAX_SNIPPETS,
     ) -> tuple[list[str], NonNegativeInt]:
         """Combine snippets from multiple sources and return total match count.
@@ -289,6 +292,11 @@ class SnippetGenerator:
 
         snippets can be empty for real in case of e.g. title match
         """
+
+        assert (
+            summary is not None or webvtt is not None
+        ), "At least one source must be present"
+
         webvtt_matches = 0
         summary_matches = 0
 
@@ -355,8 +363,8 @@ class SearchController:
                 else_=rooms.c.name,
             ).label("room_name"),
         ]
-
-        if params.query_text:
+        search_query = None
+        if params.query_text is not None:
             search_query = sqlalchemy.func.websearch_to_tsquery(
                 "english", params.query_text
             )
@@ -373,7 +381,9 @@ class SearchController:
             transcripts.join(rooms, transcripts.c.room_id == rooms.c.id, isouter=True)
         )
 
-        if params.query_text:
+        if params.query_text is not None:
+            # because already initialized based on params.query_text presence above
+            assert search_query is not None
             base_query = base_query.where(
                 transcripts.c.search_vector_en.op("@@")(search_query)
             )
@@ -393,7 +403,7 @@ class SearchController:
                 transcripts.c.source_kind == params.source_kind
             )
 
-        if params.query_text:
+        if params.query_text is not None:
             order_by = sqlalchemy.desc(sqlalchemy.text("rank"))
         else:
             order_by = sqlalchemy.desc(transcripts.c.created_at)
@@ -407,19 +417,29 @@ class SearchController:
         )
         total = await get_database().fetch_val(count_query)
 
-        def _process_result(r) -> SearchResult:
+        def _process_result(r: DbRecord) -> SearchResult:
             r_dict: Dict[str, Any] = dict(r)
+
             webvtt_raw: str | None = r_dict.pop("webvtt", None)
+            webvtt: WebVTTContent | None
             if webvtt_raw:
                 webvtt = WebVTTProcessor.parse(webvtt_raw)
             else:
                 webvtt = None
-            long_summary: str | None = r_dict.pop("long_summary", None)
+
+            long_summary_r: str | None = r_dict.pop("long_summary", None)
+            long_summary: NonEmptyString = try_parse_non_empty_string(long_summary_r)
             room_name: str | None = r_dict.pop("room_name", None)
             db_result = SearchResultDB.model_validate(r_dict)
 
-            snippets, total_match_count = SnippetGenerator.combine_sources(
-                long_summary, webvtt, params.query_text, DEFAULT_MAX_SNIPPETS
+            at_least_one_source = webvtt is not None or long_summary is not None
+            has_query = params.query_text is not None
+            snippets, total_match_count = (
+                SnippetGenerator.combine_sources(
+                    long_summary, webvtt, params.query_text, DEFAULT_MAX_SNIPPETS
+                )
+                if has_query and at_least_one_source
+                else ([], 0)
             )
 
             return SearchResult(

server/reflector/db/transcripts.py

@@ -122,6 +122,15 @@ def generate_transcript_name() -> str:
     return f"Transcript {now.strftime('%Y-%m-%d %H:%M:%S')}"
 
 
+TranscriptStatus = Literal[
+    "idle", "uploaded", "recording", "processing", "error", "ended"
+]
+
+
+class StrValue(BaseModel):
+    value: str
+
+
 class AudioWaveform(BaseModel):
     data: list[float]
 
@@ -185,7 +194,7 @@ class Transcript(BaseModel):
     id: str = Field(default_factory=generate_uuid4)
     user_id: str | None = None
     name: str = Field(default_factory=generate_transcript_name)
-    status: str = "idle"
+    status: TranscriptStatus = "idle"
     duration: float = 0
     created_at: datetime = Field(default_factory=lambda: datetime.now(timezone.utc))
     title: str | None = None
@@ -732,5 +741,27 @@ class TranscriptController:
         transcript.delete_participant(participant_id)
         await self.update(transcript, {"participants": transcript.participants_dump()})
 
+    async def set_status(
+        self, transcript_id: str, status: TranscriptStatus
+    ) -> TranscriptEvent | None:
+        """
+        Update the status of a transcript
+
+        Will add an event STATUS + update the status field of transcript
+        """
+        async with self.transaction():
+            transcript = await self.get_by_id(transcript_id)
+            if not transcript:
+                raise Exception(f"Transcript {transcript_id} not found")
+            if transcript.status == status:
+                return
+            resp = await self.append_event(
+                transcript=transcript,
+                event="STATUS",
+                data=StrValue(value=status),
+            )
+            await self.update(transcript, {"status": status})
+        return resp
+
 
 transcripts_controller = TranscriptController()

server/reflector/pipelines/main_file_pipeline.py

@@ -7,18 +7,26 @@ Uses parallel processing for transcription, diarization, and waveform generation
 """
 
 import asyncio
+import uuid
 from pathlib import Path
 
 import av
 import structlog
 from celery import shared_task
 
+from reflector.asynctask import asynctask
+from reflector.db.rooms import rooms_controller
 from reflector.db.transcripts import (
+    SourceKind,
     Transcript,
+    TranscriptStatus,
     transcripts_controller,
 )
 from reflector.logger import logger
-from reflector.pipelines.main_live_pipeline import PipelineMainBase, asynctask
+from reflector.pipelines.main_live_pipeline import (
+    PipelineMainBase,
+    broadcast_to_sockets,
+)
 from reflector.processors import (
     AudioFileWriterProcessor,
     TranscriptFinalSummaryProcessor,
@@ -43,6 +51,7 @@ from reflector.processors.types import (
 )
 from reflector.settings import settings
 from reflector.storage import get_transcripts_storage
+from reflector.worker.webhook import send_transcript_webhook
 
 
 class EmptyPipeline:
@@ -83,12 +92,27 @@ class PipelineMainFile(PipelineMainBase):
                 exc_info=result,
             )
 
+    @broadcast_to_sockets
+    async def set_status(self, transcript_id: str, status: TranscriptStatus):
+        async with self.lock_transaction():
+            return await transcripts_controller.set_status(transcript_id, status)
+
     async def process(self, file_path: Path):
         """Main entry point for file processing"""
         self.logger.info(f"Starting file pipeline for {file_path}")
 
         transcript = await self.get_transcript()
 
+        # Clear transcript as we're going to regenerate everything
+        async with self.transaction():
+            await transcripts_controller.update(
+                transcript,
+                {
+                    "events": [],
+                    "topics": [],
+                },
+            )
+
         # Extract audio and write to transcript location
         audio_path = await self.extract_and_write_audio(file_path, transcript)
 
@@ -105,6 +129,8 @@ class PipelineMainFile(PipelineMainBase):
 
         self.logger.info("File pipeline complete")
 
+        await transcripts_controller.set_status(transcript.id, "ended")
+
     async def extract_and_write_audio(
         self, file_path: Path, transcript: Transcript
     ) -> Path:
@@ -362,14 +388,34 @@ async def task_pipeline_file_process(*, transcript_id: str):
     if not transcript:
         raise Exception(f"Transcript {transcript_id} not found")
 
-    # Find the file to process
-    audio_file = next(transcript.data_path.glob("upload.*"), None)
-    if not audio_file:
-        audio_file = next(transcript.data_path.glob("audio.*"), None)
-
-    if not audio_file:
-        raise Exception("No audio file found to process")
-
-    # Run file pipeline
     pipeline = PipelineMainFile(transcript_id=transcript_id)
-    await pipeline.process(audio_file)
+    try:
+        await pipeline.set_status(transcript_id, "processing")
+
+        # Find the file to process
+        audio_file = next(transcript.data_path.glob("upload.*"), None)
+        if not audio_file:
+            audio_file = next(transcript.data_path.glob("audio.*"), None)
+
+        if not audio_file:
+            raise Exception("No audio file found to process")
+
+        await pipeline.process(audio_file)
+
+    except Exception:
+        await pipeline.set_status(transcript_id, "error")
+        raise
+
+    # Trigger webhook if this is a room recording with webhook configured
+    if transcript.source_kind == SourceKind.ROOM and transcript.room_id:
+        room = await rooms_controller.get_by_id(transcript.room_id)
+        if room and room.webhook_url:
+            logger.info(
+                "Dispatching webhook task",
+                transcript_id=transcript_id,
+                room_id=room.id,
+                webhook_url=room.webhook_url,
+            )
+            send_transcript_webhook.delay(
+                transcript_id, room.id, event_id=uuid.uuid4().hex
+            )

server/reflector/pipelines/main_live_pipeline.py

@@ -22,7 +22,7 @@ from celery import chord, current_task, group, shared_task
 from pydantic import BaseModel
 from structlog import BoundLogger as Logger
 
-from reflector.db import get_database
+from reflector.asynctask import asynctask
 from reflector.db.meetings import meeting_consent_controller, meetings_controller
 from reflector.db.recordings import recordings_controller
 from reflector.db.rooms import rooms_controller
@@ -32,6 +32,7 @@ from reflector.db.transcripts import (
     TranscriptFinalLongSummary,
     TranscriptFinalShortSummary,
     TranscriptFinalTitle,
+    TranscriptStatus,
     TranscriptText,
     TranscriptTopic,
     TranscriptWaveform,
@@ -40,8 +41,9 @@ from reflector.db.transcripts import (
 from reflector.logger import logger
 from reflector.pipelines.runner import PipelineMessage, PipelineRunner
 from reflector.processors import (
-    AudioChunkerProcessor,
+    AudioChunkerAutoProcessor,
     AudioDiarizationAutoProcessor,
+    AudioDownscaleProcessor,
     AudioFileWriterProcessor,
     AudioMergeProcessor,
     AudioTranscriptAutoProcessor,
@@ -68,29 +70,6 @@ from reflector.zulip import (
 )
 
 
-def asynctask(f):
-    @functools.wraps(f)
-    def wrapper(*args, **kwargs):
-        async def run_with_db():
-            database = get_database()
-            await database.connect()
-            try:
-                return await f(*args, **kwargs)
-            finally:
-                await database.disconnect()
-
-        coro = run_with_db()
-        try:
-            loop = asyncio.get_running_loop()
-        except RuntimeError:
-            loop = None
-        if loop and loop.is_running():
-            return loop.run_until_complete(coro)
-        return asyncio.run(coro)
-
-    return wrapper
-
-
 def broadcast_to_sockets(func):
     """
     Decorator to broadcast transcript event to websockets
@@ -187,8 +166,15 @@ class PipelineMainBase(PipelineRunner[PipelineMessage], Generic[PipelineMessage]
         ]
 
     @asynccontextmanager
-    async def transaction(self):
+    async def lock_transaction(self):
+        # This lock is to prevent multiple processor starting adding
+        # into event array at the same time
         async with self._lock:
+            yield
+
+    @asynccontextmanager
+    async def transaction(self):
+        async with self.lock_transaction():
             async with transcripts_controller.transaction():
                 yield
 
@@ -197,14 +183,14 @@ class PipelineMainBase(PipelineRunner[PipelineMessage], Generic[PipelineMessage]
         # if it's the first part, update the status of the transcript
         # but do not set the ended status yet.
         if isinstance(self, PipelineMainLive):
-            status_mapping = {
+            status_mapping: dict[str, TranscriptStatus] = {
                 "started": "recording",
                 "push": "recording",
                 "flush": "processing",
                 "error": "error",
             }
         elif isinstance(self, PipelineMainFinalSummaries):
-            status_mapping = {
+            status_mapping: dict[str, TranscriptStatus] = {
                 "push": "processing",
                 "flush": "processing",
                 "error": "error",
@@ -220,22 +206,8 @@ class PipelineMainBase(PipelineRunner[PipelineMessage], Generic[PipelineMessage]
             return
 
         # when the status of the pipeline changes, update the transcript
-        async with self.transaction():
-            transcript = await self.get_transcript()
-            if status == transcript.status:
-                return
-            resp = await transcripts_controller.append_event(
-                transcript=transcript,
-                event="STATUS",
-                data=StrValue(value=status),
-            )
-            await transcripts_controller.update(
-                transcript,
-                {
-                    "status": status,
-                },
-            )
-            return resp
+        async with self._lock:
+            return await transcripts_controller.set_status(self.transcript_id, status)
 
     @broadcast_to_sockets
     async def on_transcript(self, data):
@@ -365,7 +337,8 @@ class PipelineMainLive(PipelineMainBase):
                 path=transcript.audio_wav_filename,
                 on_duration=self.on_duration,
             ),
-            AudioChunkerProcessor(),
+            AudioDownscaleProcessor(),
+            AudioChunkerAutoProcessor(),
             AudioMergeProcessor(),
             AudioTranscriptAutoProcessor.as_threaded(),
             TranscriptLinerProcessor(),
@@ -792,7 +765,7 @@ def pipeline_post(*, transcript_id: str):
         chain_final_summaries,
     ) | task_pipeline_post_to_zulip.si(transcript_id=transcript_id)
 
-    chain.delay()
+    return chain.delay()
 
 
 @get_transcript

server/reflector/processors/__init__.py

@@ -1,5 +1,7 @@
 from .audio_chunker import AudioChunkerProcessor  # noqa: F401
+from .audio_chunker_auto import AudioChunkerAutoProcessor  # noqa: F401
 from .audio_diarization_auto import AudioDiarizationAutoProcessor  # noqa: F401
+from .audio_downscale import AudioDownscaleProcessor  # noqa: F401
 from .audio_file_writer import AudioFileWriterProcessor  # noqa: F401
 from .audio_merge import AudioMergeProcessor  # noqa: F401
 from .audio_transcript import AudioTranscriptProcessor  # noqa: F401

server/reflector/processors/audio_chunker.py

@@ -1,340 +1,78 @@
 from typing import Optional
 
 import av
-import numpy as np
-import torch
-from silero_vad import VADIterator, load_silero_vad
+from prometheus_client import Counter, Histogram
 
 from reflector.processors.base import Processor
 
 
 class AudioChunkerProcessor(Processor):
     """
-    Assemble audio frames into chunks with VAD-based speech detection
+    Base class for assembling audio frames into chunks
     """
 
     INPUT_TYPE = av.AudioFrame
     OUTPUT_TYPE = list[av.AudioFrame]
 
-    def __init__(
-        self,
-        block_frames=256,
-        max_frames=1024,
-        vad_threshold=0.5,
-        use_onnx=False,
-        min_frames=2,
-    ):
-        super().__init__()
+    m_chunk = Histogram(
+        "audio_chunker",
+        "Time spent in AudioChunker.chunk",
+        ["backend"],
+    )
+    m_chunk_call = Counter(
+        "audio_chunker_call",
+        "Number of calls to AudioChunker.chunk",
+        ["backend"],
+    )
+    m_chunk_success = Counter(
+        "audio_chunker_success",
+        "Number of successful calls to AudioChunker.chunk",
+        ["backend"],
+    )
+    m_chunk_failure = Counter(
+        "audio_chunker_failure",
+        "Number of failed calls to AudioChunker.chunk",
+        ["backend"],
+    )
+
+    def __init__(self, *args, **kwargs):
+        name = self.__class__.__name__
+        self.m_chunk = self.m_chunk.labels(name)
+        self.m_chunk_call = self.m_chunk_call.labels(name)
+        self.m_chunk_success = self.m_chunk_success.labels(name)
+        self.m_chunk_failure = self.m_chunk_failure.labels(name)
+        super().__init__(*args, **kwargs)
         self.frames: list[av.AudioFrame] = []
-        self.block_frames = block_frames
-        self.max_frames = max_frames
-        self.vad_threshold = vad_threshold
-        self.min_frames = min_frames
-
-        # Initialize Silero VAD
-        self._init_vad(use_onnx)
-
-    def _init_vad(self, use_onnx=False):
-        """Initialize Silero VAD model"""
-        try:
-            torch.set_num_threads(1)
-            self.vad_model = load_silero_vad(onnx=use_onnx)
-            self.vad_iterator = VADIterator(self.vad_model, sampling_rate=16000)
-            self.logger.info("Silero VAD initialized successfully")
-
-        except Exception as e:
-            self.logger.error(f"Failed to initialize Silero VAD: {e}")
-            self.vad_model = None
-            self.vad_iterator = None
 
     async def _push(self, data: av.AudioFrame):
-        self.frames.append(data)
-        # print("timestamp", data.pts * data.time_base * 1000)
-
-        # Check for speech segments every 32 frames (~1 second)
-        if len(self.frames) >= 32 and len(self.frames) % 32 == 0:
-            await self._process_block()
-
-        # Safety fallback - emit if we hit max frames
-        elif len(self.frames) >= self.max_frames:
-            self.logger.warning(
-                f"AudioChunkerProcessor: Reached max frames ({self.max_frames}), "
-                f"emitting first {self.max_frames // 2} frames"
-            )
-            frames_to_emit = self.frames[: self.max_frames // 2]
-            self.frames = self.frames[self.max_frames // 2 :]
-            if len(frames_to_emit) >= self.min_frames:
-                await self.emit(frames_to_emit)
-            else:
-                self.logger.debug(
-                    f"Ignoring fallback segment with {len(frames_to_emit)} frames "
-                    f"(< {self.min_frames} minimum)"
+        """Process incoming audio frame"""
+        # Validate audio format on first frame
+        if len(self.frames) == 0:
+            if data.sample_rate != 16000 or len(data.layout.channels) != 1:
+                raise ValueError(
+                    f"AudioChunkerProcessor expects 16kHz mono audio, got {data.sample_rate}Hz "
+                    f"with {len(data.layout.channels)} channel(s). "
+                    f"Use AudioDownscaleProcessor before this processor."
                 )
 
-    async def _process_block(self):
-        # Need at least 32 frames for VAD detection (~1 second)
-        if len(self.frames) < 32 or self.vad_iterator is None:
-            return
-
-        # Processing block with current buffer size
-        # print(f"Processing block: {len(self.frames)} frames in buffer")
-
         try:
-            # Convert frames to numpy array for VAD
-            audio_array = self._frames_to_numpy(self.frames)
+            self.m_chunk_call.inc()
+            with self.m_chunk.time():
+                result = await self._chunk(data)
+            self.m_chunk_success.inc()
+            if result:
+                await self.emit(result)
+        except Exception:
+            self.m_chunk_failure.inc()
+            raise
 
-            if audio_array is None:
-                # Fallback: emit all frames if conversion failed
-                frames_to_emit = self.frames[:]
-                self.frames = []
-                if len(frames_to_emit) >= self.min_frames:
-                    await self.emit(frames_to_emit)
-                else:
-                    self.logger.debug(
-                        f"Ignoring conversion-failed segment with {len(frames_to_emit)} frames "
-                        f"(< {self.min_frames} minimum)"
-                    )
-                return
-
-            # Find complete speech segments in the buffer
-            speech_end_frame = self._find_speech_segment_end(audio_array)
-
-            if speech_end_frame is None or speech_end_frame <= 0:
-                # No speech found but buffer is getting large
-                if len(self.frames) > 512:
-                    # Check if it's all silence and can be discarded
-                    # No speech segment found, buffer at {len(self.frames)} frames
-
-                    # Could emit silence or discard old frames here
-                    # For now, keep first 256 frames and discard older silence
-                    if len(self.frames) > 768:
-                        self.logger.debug(
-                            f"Discarding {len(self.frames) - 256} old frames (likely silence)"
-                        )
-                        self.frames = self.frames[-256:]
-                return
-
-            # Calculate segment timing information
-            frames_to_emit = self.frames[:speech_end_frame]
-
-            # Get timing from av.AudioFrame
-            if frames_to_emit:
-                first_frame = frames_to_emit[0]
-                last_frame = frames_to_emit[-1]
-                sample_rate = first_frame.sample_rate
-
-                # Calculate duration
-                total_samples = sum(f.samples for f in frames_to_emit)
-                duration_seconds = total_samples / sample_rate if sample_rate > 0 else 0
-
-                # Get timestamps if available
-                start_time = (
-                    first_frame.pts * first_frame.time_base if first_frame.pts else 0
-                )
-                end_time = (
-                    last_frame.pts * last_frame.time_base if last_frame.pts else 0
-                )
-
-                # Convert to HH:MM:SS format for logging
-                def format_time(seconds):
-                    if not seconds:
-                        return "00:00:00"
-                    total_seconds = int(float(seconds))
-                    hours = total_seconds // 3600
-                    minutes = (total_seconds % 3600) // 60
-                    secs = total_seconds % 60
-                    return f"{hours:02d}:{minutes:02d}:{secs:02d}"
-
-                start_formatted = format_time(start_time)
-                end_formatted = format_time(end_time)
-
-                # Keep remaining frames for next processing
-                remaining_after = len(self.frames) - speech_end_frame
-
-                # Single structured log line
-                self.logger.info(
-                    "Speech segment found",
-                    start=start_formatted,
-                    end=end_formatted,
-                    frames=speech_end_frame,
-                    duration=round(duration_seconds, 2),
-                    buffer_before=len(self.frames),
-                    remaining=remaining_after,
-                )
-
-            # Keep remaining frames for next processing
-            self.frames = self.frames[speech_end_frame:]
-
-            # Filter out segments with too few frames
-            if len(frames_to_emit) >= self.min_frames:
-                await self.emit(frames_to_emit)
-            else:
-                self.logger.debug(
-                    f"Ignoring segment with {len(frames_to_emit)} frames "
-                    f"(< {self.min_frames} minimum)"
-                )
-
-        except Exception as e:
-            self.logger.error(f"Error in VAD processing: {e}")
-            # Fallback to simple chunking
-            if len(self.frames) >= self.block_frames:
-                frames_to_emit = self.frames[: self.block_frames]
-                self.frames = self.frames[self.block_frames :]
-                if len(frames_to_emit) >= self.min_frames:
-                    await self.emit(frames_to_emit)
-                else:
-                    self.logger.debug(
-                        f"Ignoring exception-fallback segment with {len(frames_to_emit)} frames "
-                        f"(< {self.min_frames} minimum)"
-                    )
-
-    def _frames_to_numpy(self, frames: list[av.AudioFrame]) -> Optional[np.ndarray]:
-        """Convert av.AudioFrame list to numpy array for VAD processing"""
-        if not frames:
-            return None
-
-        try:
-            first_frame = frames[0]
-            original_sample_rate = first_frame.sample_rate
-
-            audio_data = []
-            for frame in frames:
-                frame_array = frame.to_ndarray()
-
-                # Handle stereo -> mono conversion
-                if len(frame_array.shape) == 2 and frame_array.shape[0] > 1:
-                    frame_array = np.mean(frame_array, axis=0)
-                elif len(frame_array.shape) == 2:
-                    frame_array = frame_array.flatten()
-
-                audio_data.append(frame_array)
-
-            if not audio_data:
-                return None
-
-            combined_audio = np.concatenate(audio_data)
-
-            # Resample from 48kHz to 16kHz if needed
-            if original_sample_rate != 16000:
-                combined_audio = self._resample_audio(
-                    combined_audio, original_sample_rate, 16000
-                )
-
-            # Ensure float32 format
-            if combined_audio.dtype == np.int16:
-                # Normalize int16 audio to float32 in range [-1.0, 1.0]
-                combined_audio = combined_audio.astype(np.float32) / 32768.0
-            elif combined_audio.dtype != np.float32:
-                combined_audio = combined_audio.astype(np.float32)
-
-            return combined_audio
-
-        except Exception as e:
-            self.logger.error(f"Error converting frames to numpy: {e}")
-
-        return None
-
-    def _resample_audio(
-        self, audio: np.ndarray, from_sr: int, to_sr: int
-    ) -> np.ndarray:
-        """Simple linear resampling from from_sr to to_sr"""
-        if from_sr == to_sr:
-            return audio
-
-        try:
-            # Simple linear interpolation resampling
-            ratio = to_sr / from_sr
-            new_length = int(len(audio) * ratio)
-
-            # Create indices for interpolation
-            old_indices = np.linspace(0, len(audio) - 1, new_length)
-            resampled = np.interp(old_indices, np.arange(len(audio)), audio)
-
-            return resampled.astype(np.float32)
-
-        except Exception as e:
-            self.logger.error("Resampling error", exc_info=e)
-            # Fallback: simple decimation/repetition
-            if from_sr > to_sr:
-                # Downsample by taking every nth sample
-                step = from_sr // to_sr
-                return audio[::step]
-            else:
-                # Upsample by repeating samples
-                repeat = to_sr // from_sr
-                return np.repeat(audio, repeat)
-
-    def _find_speech_segment_end(self, audio_array: np.ndarray) -> Optional[int]:
-        """Find complete speech segments and return frame index at segment end"""
-        if self.vad_iterator is None or len(audio_array) == 0:
-            return None
-
-        try:
-            # Process audio in 512-sample windows for VAD
-            window_size = 512
-            min_silence_windows = 3  # Require 3 windows of silence after speech
-
-            # Track speech state
-            in_speech = False
-            speech_start = None
-            speech_end = None
-            silence_count = 0
-
-            for i in range(0, len(audio_array), window_size):
-                chunk = audio_array[i : i + window_size]
-                if len(chunk) < window_size:
-                    chunk = np.pad(chunk, (0, window_size - len(chunk)))
-
-                # Detect if this window has speech
-                speech_dict = self.vad_iterator(chunk, return_seconds=True)
-
-                # VADIterator returns dict with 'start' and 'end' when speech segments are detected
-                if speech_dict:
-                    if not in_speech:
-                        # Speech started
-                        speech_start = i
-                        in_speech = True
-                        # Debug: print(f"Speech START at sample {i}, VAD: {speech_dict}")
-                    silence_count = 0  # Reset silence counter
-                    continue
-
-                if not in_speech:
-                    continue
-
-                # We're in speech but found silence
-                silence_count += 1
-                if silence_count < min_silence_windows:
-                    continue
-
-                # Found end of speech segment
-                speech_end = i - (min_silence_windows - 1) * window_size
-                # Debug: print(f"Speech END at sample {speech_end}")
-
-                # Convert sample position to frame index
-                samples_per_frame = self.frames[0].samples if self.frames else 1024
-                # Account for resampling: we process at 16kHz but frames might be 48kHz
-                resample_ratio = 48000 / 16000  # 3x
-                actual_sample_pos = int(speech_end * resample_ratio)
-                frame_index = actual_sample_pos // samples_per_frame
-
-                # Ensure we don't exceed buffer
-                frame_index = min(frame_index, len(self.frames))
-                return frame_index
-
-            return None
-
-        except Exception as e:
-            self.logger.error(f"Error finding speech segment: {e}")
-            return None
+    async def _chunk(self, data: av.AudioFrame) -> Optional[list[av.AudioFrame]]:
+        """
+        Process audio frame and return chunk when ready.
+        Subclasses should implement their chunking logic here.
+        """
+        raise NotImplementedError
 
     async def _flush(self):
-        frames = self.frames[:]
-        self.frames = []
-        if frames:
-            if len(frames) >= self.min_frames:
-                await self.emit(frames)
-            else:
-                self.logger.debug(
-                    f"Ignoring flush segment with {len(frames)} frames "
-                    f"(< {self.min_frames} minimum)"
-                )
+        """Flush any remaining frames when processing ends"""
+        raise NotImplementedError

server/reflector/processors/audio_chunker_auto.py

@@ -0,0 +1,32 @@
+import importlib
+
+from reflector.processors.audio_chunker import AudioChunkerProcessor
+from reflector.settings import settings
+
+
+class AudioChunkerAutoProcessor(AudioChunkerProcessor):
+    _registry = {}
+
+    @classmethod
+    def register(cls, name, kclass):
+        cls._registry[name] = kclass
+
+    def __new__(cls, name: str | None = None, **kwargs):
+        if name is None:
+            name = settings.AUDIO_CHUNKER_BACKEND
+        if name not in cls._registry:
+            module_name = f"reflector.processors.audio_chunker_{name}"
+            importlib.import_module(module_name)
+
+        # gather specific configuration for the processor
+        # search `AUDIO_CHUNKER_BACKEND_XXX_YYY`, push to constructor as `backend_xxx_yyy`
+        config = {}
+        name_upper = name.upper()
+        settings_prefix = "AUDIO_CHUNKER_"
+        config_prefix = f"{settings_prefix}{name_upper}_"
+        for key, value in settings:
+            if key.startswith(config_prefix):
+                config_name = key[len(settings_prefix) :].lower()
+                config[config_name] = value
+
+        return cls._registry[name](**config | kwargs)

server/reflector/processors/audio_chunker_frames.py

@@ -0,0 +1,34 @@
+from typing import Optional
+
+import av
+
+from reflector.processors.audio_chunker import AudioChunkerProcessor
+from reflector.processors.audio_chunker_auto import AudioChunkerAutoProcessor
+
+
+class AudioChunkerFramesProcessor(AudioChunkerProcessor):
+    """
+    Simple frame-based audio chunker that emits chunks after a fixed number of frames
+    """
+
+    def __init__(self, max_frames=256, **kwargs):
+        super().__init__(**kwargs)
+        self.max_frames = max_frames
+
+    async def _chunk(self, data: av.AudioFrame) -> Optional[list[av.AudioFrame]]:
+        self.frames.append(data)
+        if len(self.frames) >= self.max_frames:
+            frames_to_emit = self.frames[:]
+            self.frames = []
+            return frames_to_emit
+
+        return None
+
+    async def _flush(self):
+        frames = self.frames[:]
+        self.frames = []
+        if frames:
+            await self.emit(frames)
+
+
+AudioChunkerAutoProcessor.register("frames", AudioChunkerFramesProcessor)

server/reflector/processors/audio_chunker_silero.py

@@ -0,0 +1,298 @@
+from typing import Optional
+
+import av
+import numpy as np
+import torch
+from silero_vad import VADIterator, load_silero_vad
+
+from reflector.processors.audio_chunker import AudioChunkerProcessor
+from reflector.processors.audio_chunker_auto import AudioChunkerAutoProcessor
+
+
+class AudioChunkerSileroProcessor(AudioChunkerProcessor):
+    """
+    Assemble audio frames into chunks with VAD-based speech detection using Silero VAD
+    """
+
+    def __init__(
+        self,
+        block_frames=256,
+        max_frames=1024,
+        use_onnx=True,
+        min_frames=2,
+        **kwargs,
+    ):
+        super().__init__(**kwargs)
+        self.block_frames = block_frames
+        self.max_frames = max_frames
+        self.min_frames = min_frames
+
+        # Initialize Silero VAD
+        self._init_vad(use_onnx)
+
+    def _init_vad(self, use_onnx=False):
+        """Initialize Silero VAD model"""
+        try:
+            torch.set_num_threads(1)
+            self.vad_model = load_silero_vad(onnx=use_onnx)
+            self.vad_iterator = VADIterator(self.vad_model, sampling_rate=16000)
+            self.logger.info("Silero VAD initialized successfully")
+
+        except Exception as e:
+            self.logger.error(f"Failed to initialize Silero VAD: {e}")
+            self.vad_model = None
+            self.vad_iterator = None
+
+    async def _chunk(self, data: av.AudioFrame) -> Optional[list[av.AudioFrame]]:
+        """Process audio frame and return chunk when ready"""
+        self.frames.append(data)
+
+        # Check for speech segments every 32 frames (~1 second)
+        if len(self.frames) >= 32 and len(self.frames) % 32 == 0:
+            return await self._process_block()
+
+        # Safety fallback - emit if we hit max frames
+        elif len(self.frames) >= self.max_frames:
+            self.logger.warning(
+                f"AudioChunkerSileroProcessor: Reached max frames ({self.max_frames}), "
+                f"emitting first {self.max_frames // 2} frames"
+            )
+            frames_to_emit = self.frames[: self.max_frames // 2]
+            self.frames = self.frames[self.max_frames // 2 :]
+            if len(frames_to_emit) >= self.min_frames:
+                return frames_to_emit
+            else:
+                self.logger.debug(
+                    f"Ignoring fallback segment with {len(frames_to_emit)} frames "
+                    f"(< {self.min_frames} minimum)"
+                )
+
+        return None
+
+    async def _process_block(self) -> Optional[list[av.AudioFrame]]:
+        # Need at least 32 frames for VAD detection (~1 second)
+        if len(self.frames) < 32 or self.vad_iterator is None:
+            return None
+
+        # Processing block with current buffer size
+        print(f"Processing block: {len(self.frames)} frames in buffer")
+
+        try:
+            # Convert frames to numpy array for VAD
+            audio_array = self._frames_to_numpy(self.frames)
+
+            if audio_array is None:
+                # Fallback: emit all frames if conversion failed
+                frames_to_emit = self.frames[:]
+                self.frames = []
+                if len(frames_to_emit) >= self.min_frames:
+                    return frames_to_emit
+                else:
+                    self.logger.debug(
+                        f"Ignoring conversion-failed segment with {len(frames_to_emit)} frames "
+                        f"(< {self.min_frames} minimum)"
+                    )
+                return None
+
+            # Find complete speech segments in the buffer
+            speech_end_frame = self._find_speech_segment_end(audio_array)
+
+            if speech_end_frame is None or speech_end_frame <= 0:
+                # No speech found but buffer is getting large
+                if len(self.frames) > 512:
+                    # Check if it's all silence and can be discarded
+                    # No speech segment found, buffer at {len(self.frames)} frames
+
+                    # Could emit silence or discard old frames here
+                    # For now, keep first 256 frames and discard older silence
+                    if len(self.frames) > 768:
+                        self.logger.debug(
+                            f"Discarding {len(self.frames) - 256} old frames (likely silence)"
+                        )
+                        self.frames = self.frames[-256:]
+                return None
+
+            # Calculate segment timing information
+            frames_to_emit = self.frames[:speech_end_frame]
+
+            # Get timing from av.AudioFrame
+            if frames_to_emit:
+                first_frame = frames_to_emit[0]
+                last_frame = frames_to_emit[-1]
+                sample_rate = first_frame.sample_rate
+
+                # Calculate duration
+                total_samples = sum(f.samples for f in frames_to_emit)
+                duration_seconds = total_samples / sample_rate if sample_rate > 0 else 0
+
+                # Get timestamps if available
+                start_time = (
+                    first_frame.pts * first_frame.time_base if first_frame.pts else 0
+                )
+                end_time = (
+                    last_frame.pts * last_frame.time_base if last_frame.pts else 0
+                )
+
+                # Convert to HH:MM:SS format for logging
+                def format_time(seconds):
+                    if not seconds:
+                        return "00:00:00"
+                    total_seconds = int(float(seconds))
+                    hours = total_seconds // 3600
+                    minutes = (total_seconds % 3600) // 60
+                    secs = total_seconds % 60
+                    return f"{hours:02d}:{minutes:02d}:{secs:02d}"
+
+                start_formatted = format_time(start_time)
+                end_formatted = format_time(end_time)
+
+                # Keep remaining frames for next processing
+                remaining_after = len(self.frames) - speech_end_frame
+
+                # Single structured log line
+                self.logger.info(
+                    "Speech segment found",
+                    start=start_formatted,
+                    end=end_formatted,
+                    frames=speech_end_frame,
+                    duration=round(duration_seconds, 2),
+                    buffer_before=len(self.frames),
+                    remaining=remaining_after,
+                )
+
+            # Keep remaining frames for next processing
+            self.frames = self.frames[speech_end_frame:]
+
+            # Filter out segments with too few frames
+            if len(frames_to_emit) >= self.min_frames:
+                return frames_to_emit
+            else:
+                self.logger.debug(
+                    f"Ignoring segment with {len(frames_to_emit)} frames "
+                    f"(< {self.min_frames} minimum)"
+                )
+
+        except Exception as e:
+            self.logger.error(f"Error in VAD processing: {e}")
+            # Fallback to simple chunking
+            if len(self.frames) >= self.block_frames:
+                frames_to_emit = self.frames[: self.block_frames]
+                self.frames = self.frames[self.block_frames :]
+                if len(frames_to_emit) >= self.min_frames:
+                    return frames_to_emit
+                else:
+                    self.logger.debug(
+                        f"Ignoring exception-fallback segment with {len(frames_to_emit)} frames "
+                        f"(< {self.min_frames} minimum)"
+                    )
+
+        return None
+
+    def _frames_to_numpy(self, frames: list[av.AudioFrame]) -> Optional[np.ndarray]:
+        """Convert av.AudioFrame list to numpy array for VAD processing"""
+        if not frames:
+            return None
+
+        try:
+            audio_data = []
+            for frame in frames:
+                frame_array = frame.to_ndarray()
+
+                if len(frame_array.shape) == 2:
+                    frame_array = frame_array.flatten()
+
+                audio_data.append(frame_array)
+
+            if not audio_data:
+                return None
+
+            combined_audio = np.concatenate(audio_data)
+
+            # Ensure float32 format
+            if combined_audio.dtype == np.int16:
+                # Normalize int16 audio to float32 in range [-1.0, 1.0]
+                combined_audio = combined_audio.astype(np.float32) / 32768.0
+            elif combined_audio.dtype != np.float32:
+                combined_audio = combined_audio.astype(np.float32)
+
+            return combined_audio
+
+        except Exception as e:
+            self.logger.error(f"Error converting frames to numpy: {e}")
+
+        return None
+
+    def _find_speech_segment_end(self, audio_array: np.ndarray) -> Optional[int]:
+        """Find complete speech segments and return frame index at segment end"""
+        if self.vad_iterator is None or len(audio_array) == 0:
+            return None
+
+        try:
+            # Process audio in 512-sample windows for VAD
+            window_size = 512
+            min_silence_windows = 3  # Require 3 windows of silence after speech
+
+            # Track speech state
+            in_speech = False
+            speech_start = None
+            speech_end = None
+            silence_count = 0
+
+            for i in range(0, len(audio_array), window_size):
+                chunk = audio_array[i : i + window_size]
+                if len(chunk) < window_size:
+                    chunk = np.pad(chunk, (0, window_size - len(chunk)))
+
+                # Detect if this window has speech
+                speech_dict = self.vad_iterator(chunk, return_seconds=True)
+
+                # VADIterator returns dict with 'start' and 'end' when speech segments are detected
+                if speech_dict:
+                    if not in_speech:
+                        # Speech started
+                        speech_start = i
+                        in_speech = True
+                        # Debug: print(f"Speech START at sample {i}, VAD: {speech_dict}")
+                    silence_count = 0  # Reset silence counter
+                    continue
+
+                if not in_speech:
+                    continue
+
+                # We're in speech but found silence
+                silence_count += 1
+                if silence_count < min_silence_windows:
+                    continue
+
+                # Found end of speech segment
+                speech_end = i - (min_silence_windows - 1) * window_size
+                # Debug: print(f"Speech END at sample {speech_end}")
+
+                # Convert sample position to frame index
+                samples_per_frame = self.frames[0].samples if self.frames else 1024
+                frame_index = speech_end // samples_per_frame
+
+                # Ensure we don't exceed buffer
+                frame_index = min(frame_index, len(self.frames))
+                return frame_index
+
+            return None
+
+        except Exception as e:
+            self.logger.error(f"Error finding speech segment: {e}")
+            return None
+
+    async def _flush(self):
+        frames = self.frames[:]
+        self.frames = []
+        if frames:
+            if len(frames) >= self.min_frames:
+                await self.emit(frames)
+            else:
+                self.logger.debug(
+                    f"Ignoring flush segment with {len(frames)} frames "
+                    f"(< {self.min_frames} minimum)"
+                )
+
+
+AudioChunkerAutoProcessor.register("silero", AudioChunkerSileroProcessor)

server/reflector/processors/audio_downscale.py

@@ -0,0 +1,60 @@
+from typing import Optional
+
+import av
+from av.audio.resampler import AudioResampler
+
+from reflector.processors.base import Processor
+
+
+def copy_frame(frame: av.AudioFrame) -> av.AudioFrame:
+    frame_copy = frame.from_ndarray(
+        frame.to_ndarray(),
+        format=frame.format.name,
+        layout=frame.layout.name,
+    )
+    frame_copy.sample_rate = frame.sample_rate
+    frame_copy.pts = frame.pts
+    frame_copy.time_base = frame.time_base
+    return frame_copy
+
+
+class AudioDownscaleProcessor(Processor):
+    """
+    Downscale audio frames to 16kHz mono format
+    """
+
+    INPUT_TYPE = av.AudioFrame
+    OUTPUT_TYPE = av.AudioFrame
+
+    def __init__(self, target_rate: int = 16000, target_layout: str = "mono", **kwargs):
+        super().__init__(**kwargs)
+        self.target_rate = target_rate
+        self.target_layout = target_layout
+        self.resampler: Optional[AudioResampler] = None
+        self.needs_resampling: Optional[bool] = None
+
+    async def _push(self, data: av.AudioFrame):
+        if self.needs_resampling is None:
+            self.needs_resampling = (
+                data.sample_rate != self.target_rate
+                or data.layout.name != self.target_layout
+            )
+
+            if self.needs_resampling:
+                self.resampler = AudioResampler(
+                    format="s16", layout=self.target_layout, rate=self.target_rate
+                )
+
+        if not self.needs_resampling or not self.resampler:
+            await self.emit(data)
+            return
+
+        resampled_frames = self.resampler.resample(copy_frame(data))
+        for resampled_frame in resampled_frames:
+            await self.emit(resampled_frame)
+
+    async def _flush(self):
+        if self.needs_resampling and self.resampler:
+            final_frames = self.resampler.resample(None)
+            for frame in final_frames:
+                await self.emit(frame)

server/reflector/processors/audio_merge.py

@@ -3,24 +3,11 @@ from time import monotonic_ns
 from uuid import uuid4
 
 import av
-from av.audio.resampler import AudioResampler
 
 from reflector.processors.base import Processor
 from reflector.processors.types import AudioFile
 
 
-def copy_frame(frame: av.AudioFrame) -> av.AudioFrame:
-    frame_copy = frame.from_ndarray(
-        frame.to_ndarray(),
-        format=frame.format.name,
-        layout=frame.layout.name,
-    )
-    frame_copy.sample_rate = frame.sample_rate
-    frame_copy.pts = frame.pts
-    frame_copy.time_base = frame.time_base
-    return frame_copy
-
-
 class AudioMergeProcessor(Processor):
     """
     Merge audio frame into a single file
@@ -29,9 +16,8 @@ class AudioMergeProcessor(Processor):
     INPUT_TYPE = list[av.AudioFrame]
     OUTPUT_TYPE = AudioFile
 
-    def __init__(self, downsample_to_16k_mono: bool = True, **kwargs):
+    def __init__(self, **kwargs):
         super().__init__(**kwargs)
-        self.downsample_to_16k_mono = downsample_to_16k_mono
 
     async def _push(self, data: list[av.AudioFrame]):
         if not data:
@@ -39,72 +25,27 @@ class AudioMergeProcessor(Processor):
 
         # get audio information from first frame
         frame = data[0]
-        original_channels = len(frame.layout.channels)
-        original_sample_rate = frame.sample_rate
-        original_sample_width = frame.format.bytes
-
-        # determine if we need processing
-        needs_processing = self.downsample_to_16k_mono and (
-            original_sample_rate != 16000 or original_channels != 1
-        )
-
-        # determine output parameters
-        if self.downsample_to_16k_mono:
-            output_sample_rate = 16000
-            output_channels = 1
-            output_sample_width = 2  # 16-bit = 2 bytes
-        else:
-            output_sample_rate = original_sample_rate
-            output_channels = original_channels
-            output_sample_width = original_sample_width
+        output_channels = len(frame.layout.channels)
+        output_sample_rate = frame.sample_rate
+        output_sample_width = frame.format.bytes
 
         # create audio file
         uu = uuid4().hex
         fd = io.BytesIO()
 
-        if needs_processing:
-            # Process with PyAV resampler
-            out_container = av.open(fd, "w", format="wav")
-            out_stream = out_container.add_stream("pcm_s16le", rate=16000)
-            out_stream.layout = "mono"
+        # Use PyAV to write frames
+        out_container = av.open(fd, "w", format="wav")
+        out_stream = out_container.add_stream("pcm_s16le", rate=output_sample_rate)
+        out_stream.layout = frame.layout.name
 
-            # Create resampler if needed
-            resampler = None
-            if original_sample_rate != 16000 or original_channels != 1:
-                resampler = AudioResampler(format="s16", layout="mono", rate=16000)
-
-            for frame in data:
-                if resampler:
-                    # Resample and convert to mono
-                    # XXX for an unknown reason, if we don't use a copy of the frame, we get
-                    # Invalid Argumment from resample. Debugging indicate that when a previous processor
-                    # already used the frame (like AudioFileWriter), it make it invalid argument here.
-                    resampled_frames = resampler.resample(copy_frame(frame))
-                    for resampled_frame in resampled_frames:
-                        for packet in out_stream.encode(resampled_frame):
-                            out_container.mux(packet)
-                else:
-                    # Direct encoding without resampling
-                    for packet in out_stream.encode(frame):
-                        out_container.mux(packet)
-
-            # Flush the encoder
-            for packet in out_stream.encode(None):
+        for frame in data:
+            for packet in out_stream.encode(frame):
                 out_container.mux(packet)
-            out_container.close()
-        else:
-            # Use PyAV for original frames (no processing needed)
-            out_container = av.open(fd, "w", format="wav")
-            out_stream = out_container.add_stream("pcm_s16le", rate=output_sample_rate)
-            out_stream.layout = "mono" if output_channels == 1 else frame.layout
 
-            for frame in data:
-                for packet in out_stream.encode(frame):
-                    out_container.mux(packet)
-
-            for packet in out_stream.encode(None):
-                out_container.mux(packet)
-            out_container.close()
+        # Flush the encoder
+        for packet in out_stream.encode(None):
+            out_container.mux(packet)
+        out_container.close()
 
         fd.seek(0)
 

server/reflector/processors/audio_transcript_modal.py

@@ -12,9 +12,6 @@ API will be a POST request to TRANSCRIPT_URL:
 
 """
 
-from typing import List
-
-import aiohttp
 from openai import AsyncOpenAI
 
 from reflector.processors.audio_transcript import AudioTranscriptProcessor
@@ -25,7 +22,9 @@ from reflector.settings import settings
 
 class AudioTranscriptModalProcessor(AudioTranscriptProcessor):
     def __init__(
-        self, modal_api_key: str | None = None, batch_enabled: bool = True, **kwargs
+        self,
+        modal_api_key: str | None = None,
+        **kwargs,
     ):
         super().__init__()
         if not settings.TRANSCRIPT_URL:
@@ -35,126 +34,6 @@ class AudioTranscriptModalProcessor(AudioTranscriptProcessor):
         self.transcript_url = settings.TRANSCRIPT_URL + "/v1"
         self.timeout = settings.TRANSCRIPT_TIMEOUT
         self.modal_api_key = modal_api_key
-        self.max_batch_duration = 10.0
-        self.max_batch_files = 15
-        self.batch_enabled = batch_enabled
-        self.pending_files: List[AudioFile] = []  # Files waiting to be processed
-
-    @classmethod
-    def _calculate_duration(cls, audio_file: AudioFile) -> float:
-        """Calculate audio duration in seconds from AudioFile metadata"""
-        # Duration = total_samples / sample_rate
-        # We need to estimate total samples from the file data
-        import wave
-
-        try:
-            # Try to read as WAV file to get duration
-            audio_file.fd.seek(0)
-            with wave.open(audio_file.fd, "rb") as wav_file:
-                frames = wav_file.getnframes()
-                sample_rate = wav_file.getframerate()
-                duration = frames / sample_rate
-                return duration
-        except Exception:
-            # Fallback: estimate from file size and audio parameters
-            audio_file.fd.seek(0, 2)  # Seek to end
-            file_size = audio_file.fd.tell()
-            audio_file.fd.seek(0)  # Reset to beginning
-
-            # Estimate: file_size / (sample_rate * channels * sample_width)
-            bytes_per_second = (
-                audio_file.sample_rate
-                * audio_file.channels
-                * (audio_file.sample_width // 8)
-            )
-            estimated_duration = (
-                file_size / bytes_per_second if bytes_per_second > 0 else 0
-            )
-            return max(0, estimated_duration)
-
-    def _create_batches(self, audio_files: List[AudioFile]) -> List[List[AudioFile]]:
-        """Group audio files into batches with maximum 30s total duration"""
-        batches = []
-        current_batch = []
-        current_duration = 0.0
-
-        for audio_file in audio_files:
-            duration = self._calculate_duration(audio_file)
-
-            # If adding this file exceeds max duration, start a new batch
-            if current_duration + duration > self.max_batch_duration and current_batch:
-                batches.append(current_batch)
-                current_batch = [audio_file]
-                current_duration = duration
-            else:
-                current_batch.append(audio_file)
-                current_duration += duration
-
-        # Add the last batch if not empty
-        if current_batch:
-            batches.append(current_batch)
-
-        return batches
-
-    async def _transcript_batch(self, audio_files: List[AudioFile]) -> List[Transcript]:
-        """Transcribe a batch of audio files using the parakeet backend"""
-        if not audio_files:
-            return []
-
-        self.logger.debug(f"Batch transcribing {len(audio_files)} files")
-
-        # Prepare form data for batch request
-        data = aiohttp.FormData()
-        data.add_field("language", self.get_pref("audio:source_language", "en"))
-        data.add_field("batch", "true")
-
-        for i, audio_file in enumerate(audio_files):
-            audio_file.fd.seek(0)
-            data.add_field(
-                "files",
-                audio_file.fd,
-                filename=f"{audio_file.name}",
-                content_type="audio/wav",
-            )
-
-        # Make batch request
-        headers = {"Authorization": f"Bearer {self.modal_api_key}"}
-
-        async with aiohttp.ClientSession(
-            timeout=aiohttp.ClientTimeout(total=self.timeout)
-        ) as session:
-            async with session.post(
-                f"{self.transcript_url}/audio/transcriptions",
-                data=data,
-                headers=headers,
-            ) as response:
-                if response.status != 200:
-                    error_text = await response.text()
-                    raise Exception(
-                        f"Batch transcription failed: {response.status} {error_text}"
-                    )
-
-                result = await response.json()
-
-        # Process batch results
-        transcripts = []
-        results = result.get("results", [])
-
-        for i, (audio_file, file_result) in enumerate(zip(audio_files, results)):
-            transcript = Transcript(
-                words=[
-                    Word(
-                        text=word_info["word"],
-                        start=word_info["start"],
-                        end=word_info["end"],
-                    )
-                    for word_info in file_result.get("words", [])
-                ]
-            )
-            transcript.add_offset(audio_file.timestamp)
-            transcripts.append(transcript)
-
-        return transcripts
 
     async def _transcript(self, data: AudioFile):
         async with AsyncOpenAI(
@@ -187,96 +66,5 @@ class AudioTranscriptModalProcessor(AudioTranscriptProcessor):
 
         return transcript
 
-    async def transcript_multiple(
-        self, audio_files: List[AudioFile]
-    ) -> List[Transcript]:
-        """Transcribe multiple audio files using batching"""
-        if len(audio_files) == 1:
-            # Single file, use existing method
-            return [await self._transcript(audio_files[0])]
-
-        # Create batches with max 30s duration each
-        batches = self._create_batches(audio_files)
-
-        self.logger.debug(
-            f"Processing {len(audio_files)} files in {len(batches)} batches"
-        )
-
-        # Process all batches concurrently
-        all_transcripts = []
-
-        for batch in batches:
-            batch_transcripts = await self._transcript_batch(batch)
-            all_transcripts.extend(batch_transcripts)
-
-        return all_transcripts
-
-    async def _push(self, data: AudioFile):
-        """Override _push to support batching"""
-        if not self.batch_enabled:
-            # Use parent implementation for single file processing
-            return await super()._push(data)
-
-        # Add file to pending batch
-        self.pending_files.append(data)
-        self.logger.debug(
-            f"Added file to batch: {data.name}, batch size: {len(self.pending_files)}"
-        )
-
-        # Calculate total duration of pending files
-        total_duration = sum(self._calculate_duration(f) for f in self.pending_files)
-
-        # Process batch if it reaches max duration or has multiple files ready for optimization
-        should_process_batch = (
-            total_duration >= self.max_batch_duration
-            or len(self.pending_files) >= self.max_batch_files
-        )
-
-        if should_process_batch:
-            await self._process_pending_batch()
-
-    async def _process_pending_batch(self):
-        """Process all pending files as batches"""
-        if not self.pending_files:
-            return
-
-        self.logger.debug(f"Processing batch of {len(self.pending_files)} files")
-
-        try:
-            # Create batches respecting duration limit
-            batches = self._create_batches(self.pending_files)
-
-            # Process each batch
-            for batch in batches:
-                self.m_transcript_call.inc()
-                try:
-                    with self.m_transcript.time():
-                        # Use batch transcription
-                        transcripts = await self._transcript_batch(batch)
-
-                    self.m_transcript_success.inc()
-
-                    # Emit each transcript
-                    for transcript in transcripts:
-                        if transcript:
-                            await self.emit(transcript)
-
-                except Exception:
-                    self.m_transcript_failure.inc()
-                    raise
-                finally:
-                    # Release audio files
-                    for audio_file in batch:
-                        audio_file.release()
-
-        finally:
-            # Clear pending files
-            self.pending_files.clear()
-
-    async def _flush(self):
-        """Process any remaining files when flushing"""
-        await self._process_pending_batch()
-        await super()._flush()
-
 
 AudioTranscriptAutoProcessor.register("modal", AudioTranscriptModalProcessor)

server/reflector/processors/file_transcript_modal.py

@@ -67,6 +67,9 @@ class FileTranscriptModalProcessor(FileTranscriptProcessor):
             for word_info in result.get("words", [])
         ]
 
+        # words come not in order
+        words.sort(key=lambda w: w.start)
+
         return Transcript(words=words)
 
 

server/reflector/settings.py

@@ -1,3 +1,4 @@
+from pydantic.types import PositiveInt
 from pydantic_settings import BaseSettings, SettingsConfigDict
 
 
@@ -21,6 +22,10 @@ class Settings(BaseSettings):
     # local data directory
     DATA_DIR: str = "./data"
 
+    # Audio Chunking
+    # backends: silero, frames
+    AUDIO_CHUNKER_BACKEND: str = "frames"
+
     # Audio Transcription
     # backends: whisper, modal
     TRANSCRIPT_BACKEND: str = "whisper"
@@ -86,9 +91,8 @@ class Settings(BaseSettings):
     AUTH_JWT_PUBLIC_KEY: str | None = "authentik.monadical.com_public.pem"
     AUTH_JWT_AUDIENCE: str | None = None
 
-    # API public mode
-    # if set, all anonymous record will be public
     PUBLIC_MODE: bool = False
+    PUBLIC_DATA_RETENTION_DAYS: PositiveInt = 7
 
     # Min transcript length to generate topic + summary
     MIN_TRANSCRIPT_LENGTH: int = 750

server/reflector/tools/cleanup_old_data.py

@@ -0,0 +1,72 @@
+#!/usr/bin/env python
+"""
+Manual cleanup tool for old public data.
+Uses the same implementation as the Celery worker task.
+"""
+
+import argparse
+import asyncio
+import sys
+
+import structlog
+
+from reflector.settings import settings
+from reflector.worker.cleanup import _cleanup_old_public_data
+
+logger = structlog.get_logger(__name__)
+
+
+async def cleanup_old_data(days: int = 7):
+    logger.info(
+        "Starting manual cleanup",
+        retention_days=days,
+        public_mode=settings.PUBLIC_MODE,
+    )
+
+    if not settings.PUBLIC_MODE:
+        logger.critical(
+            "WARNING: PUBLIC_MODE is False. "
+            "This tool is intended for public instances only."
+        )
+        raise Exception("Tool intended for public instances only")
+
+    result = await _cleanup_old_public_data(days=days)
+
+    if result:
+        logger.info(
+            "Cleanup completed",
+            transcripts_deleted=result.get("transcripts_deleted", 0),
+            meetings_deleted=result.get("meetings_deleted", 0),
+            recordings_deleted=result.get("recordings_deleted", 0),
+            errors_count=len(result.get("errors", [])),
+        )
+        if result.get("errors"):
+            logger.warning(
+                "Errors encountered during cleanup:", errors=result["errors"][:10]
+            )
+    else:
+        logger.info("Cleanup skipped or completed without results")
+
+
+def main():
+    parser = argparse.ArgumentParser(
+        description="Clean up old transcripts and meetings"
+    )
+    parser.add_argument(
+        "--days",
+        type=int,
+        default=7,
+        help="Number of days to keep data (default: 7)",
+    )
+
+    args = parser.parse_args()
+
+    if args.days < 1:
+        logger.error("Days must be at least 1")
+        sys.exit(1)
+
+    asyncio.run(cleanup_old_data(days=args.days))
+
+
+if __name__ == "__main__":
+    main()

server/reflector/tools/process.py

@@ -1,292 +1,204 @@
 """
 Process audio file with diarization support
-===========================================
-
-Extended version of process.py that includes speaker diarization.
-This tool processes audio files locally without requiring the full server infrastructure.
 """
 
+import argparse
 import asyncio
-import tempfile
-import uuid
+import json
+import shutil
+import sys
+import time
 from pathlib import Path
-from typing import List
-
-import av
+from typing import Any, Dict, List, Literal
 
+from reflector.db.transcripts import SourceKind, TranscriptTopic, transcripts_controller
 from reflector.logger import logger
-from reflector.processors import (
-    AudioChunkerProcessor,
-    AudioFileWriterProcessor,
-    AudioMergeProcessor,
-    AudioTranscriptAutoProcessor,
-    Pipeline,
-    PipelineEvent,
-    TranscriptFinalSummaryProcessor,
-    TranscriptFinalTitleProcessor,
-    TranscriptLinerProcessor,
-    TranscriptTopicDetectorProcessor,
-    TranscriptTranslatorAutoProcessor,
+from reflector.pipelines.main_file_pipeline import (
+    task_pipeline_file_process as task_pipeline_file_process,
 )
-from reflector.processors.base import BroadcastProcessor, Processor
-from reflector.processors.types import (
-    AudioDiarizationInput,
-    TitleSummary,
-    TitleSummaryWithId,
+from reflector.pipelines.main_live_pipeline import pipeline_post as live_pipeline_post
+from reflector.pipelines.main_live_pipeline import (
+    pipeline_process as live_pipeline_process,
 )
 
 
-class TopicCollectorProcessor(Processor):
-    """Collect topics for diarization"""
+def serialize_topics(topics: List[TranscriptTopic]) -> List[Dict[str, Any]]:
+    """Convert TranscriptTopic objects to JSON-serializable dicts"""
+    serialized = []
+    for topic in topics:
+        topic_dict = topic.model_dump()
+        serialized.append(topic_dict)
+    return serialized
 
-    INPUT_TYPE = TitleSummary
-    OUTPUT_TYPE = TitleSummary
 
-    def __init__(self, **kwargs):
-        super().__init__(**kwargs)
-        self.topics: List[TitleSummaryWithId] = []
-        self._topic_id = 0
+def debug_print_speakers(serialized_topics: List[Dict[str, Any]]) -> None:
+    """Print debug info about speakers found in topics"""
+    all_speakers = set()
+    for topic_dict in serialized_topics:
+        for word in topic_dict.get("words", []):
+            all_speakers.add(word.get("speaker", 0))
 
-    async def _push(self, data: TitleSummary):
-        # Convert to TitleSummaryWithId and collect
-        self._topic_id += 1
-        topic_with_id = TitleSummaryWithId(
-            id=str(self._topic_id),
-            title=data.title,
-            summary=data.summary,
-            timestamp=data.timestamp,
-            duration=data.duration,
-            transcript=data.transcript,
+    print(
+        f"Found {len(serialized_topics)} topics with speakers: {all_speakers}",
+        file=sys.stderr,
+    )
+
+
+TranscriptId = str
+
+
+# common interface for every flow: it needs an Entry in db with specific ceremony (file path + status + actual file in file system)
+# ideally we want to get rid of it at some point
+async def prepare_entry(
+    source_path: str,
+    source_language: str,
+    target_language: str,
+) -> TranscriptId:
+    file_path = Path(source_path)
+
+    transcript = await transcripts_controller.add(
+        file_path.name,
+        # note that the real file upload has SourceKind: LIVE for the reason of it's an error
+        source_kind=SourceKind.FILE,
+        source_language=source_language,
+        target_language=target_language,
+        user_id=None,
+    )
+
+    logger.info(
+        f"Created empty transcript {transcript.id} for file {file_path.name} because technically we need an empty transcript before we start transcript"
+    )
+
+    # pipelines expect files as upload.*
+
+    extension = file_path.suffix
+    upload_path = transcript.data_path / f"upload{extension}"
+    upload_path.parent.mkdir(parents=True, exist_ok=True)
+    shutil.copy2(source_path, upload_path)
+    logger.info(f"Copied {source_path} to {upload_path}")
+
+    # pipelines expect entity status "uploaded"
+    await transcripts_controller.update(transcript, {"status": "uploaded"})
+
+    return transcript.id
+
+
+# same reason as prepare_entry
+async def extract_result_from_entry(
+    transcript_id: TranscriptId, output_path: str
+) -> None:
+    post_final_transcript = await transcripts_controller.get_by_id(transcript_id)
+
+    # assert post_final_transcript.status == "ended"
+    # File pipeline doesn't set status to "ended", only live pipeline does https://github.com/Monadical-SAS/reflector/issues/582
+    topics = post_final_transcript.topics
+    if not topics:
+        raise RuntimeError(
+            f"No topics found for transcript {transcript_id} after processing"
         )
-        self.topics.append(topic_with_id)
 
-        # Pass through the original topic
-        await self.emit(data)
+    serialized_topics = serialize_topics(topics)
 
-    def get_topics(self) -> List[TitleSummaryWithId]:
-        return self.topics
+    if output_path:
+        # Write to JSON file
+        with open(output_path, "w") as f:
+            for topic_dict in serialized_topics:
+                json.dump(topic_dict, f)
+                f.write("\n")
+        print(f"Results written to {output_path}", file=sys.stderr)
+    else:
+        # Write to stdout as JSONL
+        for topic_dict in serialized_topics:
+            print(json.dumps(topic_dict))
+
+    debug_print_speakers(serialized_topics)
 
 
-async def process_audio_file(
-    filename,
-    event_callback,
-    only_transcript=False,
-    source_language="en",
-    target_language="en",
-    enable_diarization=True,
-    diarization_backend="pyannote",
+async def process_live_pipeline(
+    transcript_id: TranscriptId,
 ):
-    # Create temp file for audio if diarization is enabled
-    audio_temp_path = None
-    if enable_diarization:
-        audio_temp_file = tempfile.NamedTemporaryFile(suffix=".wav", delete=False)
-        audio_temp_path = audio_temp_file.name
-        audio_temp_file.close()
+    """Process transcript_id with transcription and diarization"""
 
-    # Create processor for collecting topics
-    topic_collector = TopicCollectorProcessor()
+    print(f"Processing transcript_id {transcript_id}...", file=sys.stderr)
+    await live_pipeline_process(transcript_id=transcript_id)
+    print(f"Processing complete for transcript {transcript_id}", file=sys.stderr)
 
-    # Build pipeline for audio processing
-    processors = []
+    pre_final_transcript = await transcripts_controller.get_by_id(transcript_id)
 
-    # Add audio file writer at the beginning if diarization is enabled
-    if enable_diarization:
-        processors.append(AudioFileWriterProcessor(audio_temp_path))
+    # assert documented behaviour: after process, the pipeline isn't ended. this is the reason of calling pipeline_post
+    assert pre_final_transcript.status != "ended"
 
-    # Add the rest of the processors
-    processors += [
-        AudioChunkerProcessor(),
-        AudioMergeProcessor(),
-        AudioTranscriptAutoProcessor.as_threaded(),
-        TranscriptLinerProcessor(),
-        TranscriptTranslatorAutoProcessor.as_threaded(),
-    ]
+    # at this point, diarization is running but we have no access to it. run diarization in parallel - one will hopefully win after polling
+    result = live_pipeline_post(transcript_id=transcript_id)
 
-    if not only_transcript:
-        processors += [
-            TranscriptTopicDetectorProcessor.as_threaded(),
-            # Collect topics for diarization
-            topic_collector,
-            BroadcastProcessor(
-                processors=[
-                    TranscriptFinalTitleProcessor.as_threaded(),
-                    TranscriptFinalSummaryProcessor.as_threaded(),
-                ],
-            ),
-        ]
-
-    # Create main pipeline
-    pipeline = Pipeline(*processors)
-    pipeline.set_pref("audio:source_language", source_language)
-    pipeline.set_pref("audio:target_language", target_language)
-    pipeline.describe()
-    pipeline.on(event_callback)
-
-    # Start processing audio
-    logger.info(f"Opening {filename}")
-    container = av.open(filename)
-    try:
-        logger.info("Start pushing audio into the pipeline")
-        for frame in container.decode(audio=0):
-            await pipeline.push(frame)
-    finally:
-        logger.info("Flushing the pipeline")
-        await pipeline.flush()
-
-    # Run diarization if enabled and we have topics
-    if enable_diarization and not only_transcript and audio_temp_path:
-        topics = topic_collector.get_topics()
-
-        if topics:
-            logger.info(f"Starting diarization with {len(topics)} topics")
-
-            try:
-                from reflector.processors import AudioDiarizationAutoProcessor
-
-                diarization_processor = AudioDiarizationAutoProcessor(
-                    name=diarization_backend
-                )
-
-                diarization_processor.set_pipeline(pipeline)
-
-                # For Modal backend, we need to upload the file to S3 first
-                if diarization_backend == "modal":
-                    from datetime import datetime
-
-                    from reflector.storage import get_transcripts_storage
-                    from reflector.utils.s3_temp_file import S3TemporaryFile
-
-                    storage = get_transcripts_storage()
-
-                    # Generate a unique filename in evaluation folder
-                    timestamp = datetime.utcnow().strftime("%Y%m%d_%H%M%S")
-                    audio_filename = f"evaluation/diarization_temp/{timestamp}_{uuid.uuid4().hex}.wav"
-
-                    # Use context manager for automatic cleanup
-                    async with S3TemporaryFile(storage, audio_filename) as s3_file:
-                        # Read and upload the audio file
-                        with open(audio_temp_path, "rb") as f:
-                            audio_data = f.read()
-
-                        audio_url = await s3_file.upload(audio_data)
-                        logger.info(f"Uploaded audio to S3: {audio_filename}")
-
-                        # Create diarization input with S3 URL
-                        diarization_input = AudioDiarizationInput(
-                            audio_url=audio_url, topics=topics
-                        )
-
-                        # Run diarization
-                        await diarization_processor.push(diarization_input)
-                        await diarization_processor.flush()
-
-                        logger.info("Diarization complete")
-                        # File will be automatically cleaned up when exiting the context
-                else:
-                    # For local backend, use local file path
-                    audio_url = audio_temp_path
-
-                    # Create diarization input
-                    diarization_input = AudioDiarizationInput(
-                        audio_url=audio_url, topics=topics
-                    )
-
-                    # Run diarization
-                    await diarization_processor.push(diarization_input)
-                    await diarization_processor.flush()
-
-                    logger.info("Diarization complete")
-
-            except ImportError as e:
-                logger.error(f"Failed to import diarization dependencies: {e}")
-                logger.error(
-                    "Install with: uv pip install pyannote.audio torch torchaudio"
-                )
-                logger.error(
-                    "And set HF_TOKEN environment variable for pyannote models"
-                )
-                raise SystemExit(1)
-            except Exception as e:
-                logger.error(f"Diarization failed: {e}")
-                raise SystemExit(1)
-        else:
-            logger.warning("Skipping diarization: no topics available")
-
-    # Clean up temp file
-    if audio_temp_path:
-        try:
-            Path(audio_temp_path).unlink()
-        except Exception as e:
-            logger.warning(f"Failed to clean up temp file {audio_temp_path}: {e}")
-
-    logger.info("All done!")
+    # result.ready() blocks even without await; it mutates result also
+    while not result.ready():
+        print(f"Status: {result.state}")
+        time.sleep(2)
 
 
 async def process_file_pipeline(
-    filename: str,
-    event_callback,
-    source_language="en",
-    target_language="en",
-    enable_diarization=True,
-    diarization_backend="modal",
+    transcript_id: TranscriptId,
 ):
     """Process audio/video file using the optimized file pipeline"""
+
+    # task_pipeline_file_process is a Celery task, need to use .delay() for async execution
+    result = task_pipeline_file_process.delay(transcript_id=transcript_id)
+
+    # Wait for the Celery task to complete
+    while not result.ready():
+        print(f"File pipeline status: {result.state}", file=sys.stderr)
+        time.sleep(2)
+
+    logger.info("File pipeline processing complete")
+
+
+async def process(
+    source_path: str,
+    source_language: str,
+    target_language: str,
+    pipeline: Literal["live", "file"],
+    output_path: str = None,
+):
+    from reflector.db import get_database
+
+    database = get_database()
+    # db connect is a part of ceremony
+    await database.connect()
+
     try:
-        from reflector.db import database
-        from reflector.db.transcripts import SourceKind, transcripts_controller
-        from reflector.pipelines.main_file_pipeline import PipelineMainFile
-
-        await database.connect()
-        try:
-            # Create a temporary transcript for processing
-            transcript = await transcripts_controller.add(
-                "",
-                source_kind=SourceKind.FILE,
-                source_language=source_language,
-                target_language=target_language,
-            )
-
-            # Process the file
-            pipeline = PipelineMainFile(transcript_id=transcript.id)
-            await pipeline.process(Path(filename))
-
-            logger.info("File pipeline processing complete")
-
-        finally:
-            await database.disconnect()
-    except ImportError as e:
-        logger.error(f"File pipeline not available: {e}")
-        logger.info("Falling back to stream pipeline")
-        # Fall back to stream pipeline
-        await process_audio_file(
-            filename,
-            event_callback,
-            only_transcript=False,
-            source_language=source_language,
-            target_language=target_language,
-            enable_diarization=enable_diarization,
-            diarization_backend=diarization_backend,
+        transcript_id = await prepare_entry(
+            source_path,
+            source_language,
+            target_language,
         )
 
+        pipeline_handlers = {
+            "live": process_live_pipeline,
+            "file": process_file_pipeline,
+        }
+
+        handler = pipeline_handlers.get(pipeline)
+        if not handler:
+            raise ValueError(f"Unknown pipeline type: {pipeline}")
+
+        await handler(transcript_id)
+
+        await extract_result_from_entry(transcript_id, output_path)
+    finally:
+        await database.disconnect()
+
 
 if __name__ == "__main__":
-    import argparse
-    import os
-
     parser = argparse.ArgumentParser(
-        description="Process audio files with optional speaker diarization"
+        description="Process audio files with speaker diarization"
     )
     parser.add_argument("source", help="Source file (mp3, wav, mp4...)")
     parser.add_argument(
-        "--stream",
-        action="store_true",
-        help="Use streaming pipeline (original frame-based processing)",
-    )
-    parser.add_argument(
-        "--only-transcript",
-        "-t",
-        action="store_true",
-        help="Only generate transcript without topics/summaries",
+        "--pipeline",
+        required=True,
+        choices=["live", "file"],
+        help="Pipeline type to use for processing (live: streaming/incremental, file: batch/parallel)",
     )
     parser.add_argument(
         "--source-language", default="en", help="Source language code (default: en)"
@@ -295,81 +207,14 @@ if __name__ == "__main__":
         "--target-language", default="en", help="Target language code (default: en)"
     )
     parser.add_argument("--output", "-o", help="Output file (output.jsonl)")
-    parser.add_argument(
-        "--enable-diarization",
-        "-d",
-        action="store_true",
-        help="Enable speaker diarization",
-    )
-    parser.add_argument(
-        "--diarization-backend",
-        default="pyannote",
-        choices=["pyannote", "modal"],
-        help="Diarization backend to use (default: pyannote)",
-    )
     args = parser.parse_args()
 
-    if "REDIS_HOST" not in os.environ:
-        os.environ["REDIS_HOST"] = "localhost"
-
-    output_fd = None
-    if args.output:
-        output_fd = open(args.output, "w")
-
-    async def event_callback(event: PipelineEvent):
-        processor = event.processor
-        data = event.data
-
-        # Ignore internal processors
-        if processor in (
-            "AudioChunkerProcessor",
-            "AudioMergeProcessor",
-            "AudioFileWriterProcessor",
-            "TopicCollectorProcessor",
-            "BroadcastProcessor",
-        ):
-            return
-
-        # If diarization is enabled, skip the original topic events from the pipeline
-        # The diarization processor will emit the same topics but with speaker info
-        if processor == "TranscriptTopicDetectorProcessor" and args.enable_diarization:
-            return
-
-        # Log all events
-        logger.info(f"Event: {processor} - {type(data).__name__}")
-
-        # Write to output
-        if output_fd:
-            output_fd.write(event.model_dump_json())
-            output_fd.write("\n")
-            output_fd.flush()
-
-    if args.stream:
-        # Use original streaming pipeline
-        asyncio.run(
-            process_audio_file(
-                args.source,
-                event_callback,
-                only_transcript=args.only_transcript,
-                source_language=args.source_language,
-                target_language=args.target_language,
-                enable_diarization=args.enable_diarization,
-                diarization_backend=args.diarization_backend,
-            )
+    asyncio.run(
+        process(
+            args.source,
+            args.source_language,
+            args.target_language,
+            args.pipeline,
+            args.output,
         )
-    else:
-        # Use optimized file pipeline (default)
-        asyncio.run(
-            process_file_pipeline(
-                args.source,
-                event_callback,
-                source_language=args.source_language,
-                target_language=args.target_language,
-                enable_diarization=args.enable_diarization,
-                diarization_backend=args.diarization_backend,
-            )
-        )
-
-    if output_fd:
-        output_fd.close()
-        logger.info(f"Output written to {args.output}")
+    )

server/reflector/tools/process_with_diarization.py

@@ -1,315 +0,0 @@
-"""
-@vibe-generated
-Process audio file with diarization support
-===========================================
-
-Extended version of process.py that includes speaker diarization.
-This tool processes audio files locally without requiring the full server infrastructure.
-"""
-
-import asyncio
-import tempfile
-import uuid
-from pathlib import Path
-from typing import List
-
-import av
-
-from reflector.logger import logger
-from reflector.processors import (
-    AudioChunkerProcessor,
-    AudioFileWriterProcessor,
-    AudioMergeProcessor,
-    AudioTranscriptAutoProcessor,
-    Pipeline,
-    PipelineEvent,
-    TranscriptFinalSummaryProcessor,
-    TranscriptFinalTitleProcessor,
-    TranscriptLinerProcessor,
-    TranscriptTopicDetectorProcessor,
-    TranscriptTranslatorAutoProcessor,
-)
-from reflector.processors.base import BroadcastProcessor, Processor
-from reflector.processors.types import (
-    AudioDiarizationInput,
-    TitleSummary,
-    TitleSummaryWithId,
-)
-
-
-class TopicCollectorProcessor(Processor):
-    """Collect topics for diarization"""
-
-    INPUT_TYPE = TitleSummary
-    OUTPUT_TYPE = TitleSummary
-
-    def __init__(self, **kwargs):
-        super().__init__(**kwargs)
-        self.topics: List[TitleSummaryWithId] = []
-        self._topic_id = 0
-
-    async def _push(self, data: TitleSummary):
-        # Convert to TitleSummaryWithId and collect
-        self._topic_id += 1
-        topic_with_id = TitleSummaryWithId(
-            id=str(self._topic_id),
-            title=data.title,
-            summary=data.summary,
-            timestamp=data.timestamp,
-            duration=data.duration,
-            transcript=data.transcript,
-        )
-        self.topics.append(topic_with_id)
-
-        # Pass through the original topic
-        await self.emit(data)
-
-    def get_topics(self) -> List[TitleSummaryWithId]:
-        return self.topics
-
-
-async def process_audio_file_with_diarization(
-    filename,
-    event_callback,
-    only_transcript=False,
-    source_language="en",
-    target_language="en",
-    enable_diarization=True,
-    diarization_backend="modal",
-):
-    # Create temp file for audio if diarization is enabled
-    audio_temp_path = None
-    if enable_diarization:
-        audio_temp_file = tempfile.NamedTemporaryFile(suffix=".wav", delete=False)
-        audio_temp_path = audio_temp_file.name
-        audio_temp_file.close()
-
-    # Create processor for collecting topics
-    topic_collector = TopicCollectorProcessor()
-
-    # Build pipeline for audio processing
-    processors = []
-
-    # Add audio file writer at the beginning if diarization is enabled
-    if enable_diarization:
-        processors.append(AudioFileWriterProcessor(audio_temp_path))
-
-    # Add the rest of the processors
-    processors += [
-        AudioChunkerProcessor(),
-        AudioMergeProcessor(),
-        AudioTranscriptAutoProcessor.as_threaded(),
-    ]
-
-    processors += [
-        TranscriptLinerProcessor(),
-        TranscriptTranslatorAutoProcessor.as_threaded(),
-    ]
-
-    if not only_transcript:
-        processors += [
-            TranscriptTopicDetectorProcessor.as_threaded(),
-            # Collect topics for diarization
-            topic_collector,
-            BroadcastProcessor(
-                processors=[
-                    TranscriptFinalTitleProcessor.as_threaded(),
-                    TranscriptFinalSummaryProcessor.as_threaded(),
-                ],
-            ),
-        ]
-
-    # Create main pipeline
-    pipeline = Pipeline(*processors)
-    pipeline.set_pref("audio:source_language", source_language)
-    pipeline.set_pref("audio:target_language", target_language)
-    pipeline.describe()
-    pipeline.on(event_callback)
-
-    # Start processing audio
-    logger.info(f"Opening {filename}")
-    container = av.open(filename)
-    try:
-        logger.info("Start pushing audio into the pipeline")
-        for frame in container.decode(audio=0):
-            await pipeline.push(frame)
-    finally:
-        logger.info("Flushing the pipeline")
-        await pipeline.flush()
-
-    # Run diarization if enabled and we have topics
-    if enable_diarization and not only_transcript and audio_temp_path:
-        topics = topic_collector.get_topics()
-
-        if topics:
-            logger.info(f"Starting diarization with {len(topics)} topics")
-
-            try:
-                from reflector.processors import AudioDiarizationAutoProcessor
-
-                diarization_processor = AudioDiarizationAutoProcessor(
-                    name=diarization_backend
-                )
-
-                diarization_processor.set_pipeline(pipeline)
-
-                # For Modal backend, we need to upload the file to S3 first
-                if diarization_backend == "modal":
-                    from datetime import datetime, timezone
-
-                    from reflector.storage import get_transcripts_storage
-                    from reflector.utils.s3_temp_file import S3TemporaryFile
-
-                    storage = get_transcripts_storage()
-
-                    # Generate a unique filename in evaluation folder
-                    timestamp = datetime.now(timezone.utc).strftime("%Y%m%d_%H%M%S")
-                    audio_filename = f"evaluation/diarization_temp/{timestamp}_{uuid.uuid4().hex}.wav"
-
-                    # Use context manager for automatic cleanup
-                    async with S3TemporaryFile(storage, audio_filename) as s3_file:
-                        # Read and upload the audio file
-                        with open(audio_temp_path, "rb") as f:
-                            audio_data = f.read()
-
-                        audio_url = await s3_file.upload(audio_data)
-                        logger.info(f"Uploaded audio to S3: {audio_filename}")
-
-                        # Create diarization input with S3 URL
-                        diarization_input = AudioDiarizationInput(
-                            audio_url=audio_url, topics=topics
-                        )
-
-                        # Run diarization
-                        await diarization_processor.push(diarization_input)
-                        await diarization_processor.flush()
-
-                        logger.info("Diarization complete")
-                        # File will be automatically cleaned up when exiting the context
-                else:
-                    # For local backend, use local file path
-                    audio_url = audio_temp_path
-
-                    # Create diarization input
-                    diarization_input = AudioDiarizationInput(
-                        audio_url=audio_url, topics=topics
-                    )
-
-                    # Run diarization
-                    await diarization_processor.push(diarization_input)
-                    await diarization_processor.flush()
-
-                    logger.info("Diarization complete")
-
-            except ImportError as e:
-                logger.error(f"Failed to import diarization dependencies: {e}")
-                logger.error(
-                    "Install with: uv pip install pyannote.audio torch torchaudio"
-                )
-                logger.error(
-                    "And set HF_TOKEN environment variable for pyannote models"
-                )
-                raise SystemExit(1)
-            except Exception as e:
-                logger.error(f"Diarization failed: {e}")
-                raise SystemExit(1)
-        else:
-            logger.warning("Skipping diarization: no topics available")
-
-    # Clean up temp file
-    if audio_temp_path:
-        try:
-            Path(audio_temp_path).unlink()
-        except Exception as e:
-            logger.warning(f"Failed to clean up temp file {audio_temp_path}: {e}")
-
-    logger.info("All done!")
-
-
-if __name__ == "__main__":
-    import argparse
-    import os
-
-    parser = argparse.ArgumentParser(
-        description="Process audio files with optional speaker diarization"
-    )
-    parser.add_argument("source", help="Source file (mp3, wav, mp4...)")
-    parser.add_argument(
-        "--only-transcript",
-        "-t",
-        action="store_true",
-        help="Only generate transcript without topics/summaries",
-    )
-    parser.add_argument(
-        "--source-language", default="en", help="Source language code (default: en)"
-    )
-    parser.add_argument(
-        "--target-language", default="en", help="Target language code (default: en)"
-    )
-    parser.add_argument("--output", "-o", help="Output file (output.jsonl)")
-    parser.add_argument(
-        "--enable-diarization",
-        "-d",
-        action="store_true",
-        help="Enable speaker diarization",
-    )
-    parser.add_argument(
-        "--diarization-backend",
-        default="modal",
-        choices=["modal"],
-        help="Diarization backend to use (default: modal)",
-    )
-    args = parser.parse_args()
-
-    # Set REDIS_HOST to localhost if not provided
-    if "REDIS_HOST" not in os.environ:
-        os.environ["REDIS_HOST"] = "localhost"
-        logger.info("REDIS_HOST not set, defaulting to localhost")
-
-    output_fd = None
-    if args.output:
-        output_fd = open(args.output, "w")
-
-    async def event_callback(event: PipelineEvent):
-        processor = event.processor
-        data = event.data
-
-        # Ignore internal processors
-        if processor in (
-            "AudioChunkerProcessor",
-            "AudioMergeProcessor",
-            "AudioFileWriterProcessor",
-            "TopicCollectorProcessor",
-            "BroadcastProcessor",
-        ):
-            return
-
-        # If diarization is enabled, skip the original topic events from the pipeline
-        # The diarization processor will emit the same topics but with speaker info
-        if processor == "TranscriptTopicDetectorProcessor" and args.enable_diarization:
-            return
-
-        # Log all events
-        logger.info(f"Event: {processor} - {type(data).__name__}")
-
-        # Write to output
-        if output_fd:
-            output_fd.write(event.model_dump_json())
-            output_fd.write("\n")
-            output_fd.flush()
-
-    asyncio.run(
-        process_audio_file_with_diarization(
-            args.source,
-            event_callback,
-            only_transcript=args.only_transcript,
-            source_language=args.source_language,
-            target_language=args.target_language,
-            enable_diarization=args.enable_diarization,
-            diarization_backend=args.diarization_backend,
-        )
-    )
-
-    if output_fd:
-        output_fd.close()
-        logger.info(f"Output written to {args.output}")

server/reflector/tools/runpipeline.py

@@ -53,7 +53,7 @@ async def run_single_processor(args):
     async def event_callback(event: PipelineEvent):
         processor = event.processor
         # ignore some processor
-        if processor in ("AudioChunkerProcessor", "AudioMergeProcessor"):
+        if processor in ("AudioChunkerAutoProcessor", "AudioMergeProcessor"):
             return
         print(f"Event: {event}")
         if output_fd:

server/reflector/tools/test_diarization.py

@@ -1,96 +0,0 @@
-#!/usr/bin/env python3
-"""
-@vibe-generated
-Test script for the diarization CLI tool
-=========================================
-
-This script helps test the diarization functionality with sample audio files.
-"""
-
-import asyncio
-import sys
-from pathlib import Path
-
-from reflector.logger import logger
-
-
-async def test_diarization(audio_file: str):
-    """Test the diarization functionality"""
-
-    # Import the processing function
-    from process_with_diarization import process_audio_file_with_diarization
-
-    # Collect events
-    events = []
-
-    async def event_callback(event):
-        events.append({"processor": event.processor, "data": event.data})
-        logger.info(f"Event from {event.processor}")
-
-    # Process the audio file
-    logger.info(f"Processing audio file: {audio_file}")
-
-    try:
-        await process_audio_file_with_diarization(
-            audio_file,
-            event_callback,
-            only_transcript=False,
-            source_language="en",
-            target_language="en",
-            enable_diarization=True,
-            diarization_backend="modal",
-        )
-
-        # Analyze results
-        logger.info(f"Processing complete. Received {len(events)} events")
-
-        # Look for diarization results
-        diarized_topics = []
-        for event in events:
-            if "TitleSummary" in event["processor"]:
-                # Check if words have speaker information
-                if hasattr(event["data"], "transcript") and event["data"].transcript:
-                    words = event["data"].transcript.words
-                    if words and hasattr(words[0], "speaker"):
-                        speakers = set(
-                            w.speaker for w in words if hasattr(w, "speaker")
-                        )
-                        logger.info(
-                            f"Found {len(speakers)} speakers in topic: {event['data'].title}"
-                        )
-                        diarized_topics.append(event["data"])
-
-        if diarized_topics:
-            logger.info(f"Successfully diarized {len(diarized_topics)} topics")
-
-            # Print sample output
-            sample_topic = diarized_topics[0]
-            logger.info("Sample diarized output:")
-            for i, word in enumerate(sample_topic.transcript.words[:10]):
-                logger.info(f"  Word {i}: '{word.text}' - Speaker {word.speaker}")
-        else:
-            logger.warning("No diarization results found in output")
-
-        return events
-
-    except Exception as e:
-        logger.error(f"Error during processing: {e}")
-        raise
-
-
-def main():
-    if len(sys.argv) < 2:
-        print("Usage: python test_diarization.py <audio_file>")
-        sys.exit(1)
-
-    audio_file = sys.argv[1]
-    if not Path(audio_file).exists():
-        print(f"Error: Audio file '{audio_file}' not found")
-        sys.exit(1)
-
-    # Run the test
-    asyncio.run(test_diarization(audio_file))
-
-
-if __name__ == "__main__":
-    main()

server/reflector/utils/string.py

@@ -0,0 +1,20 @@
+from typing import Annotated
+
+from pydantic import Field, TypeAdapter, constr
+
+NonEmptyStringBase = constr(min_length=1, strip_whitespace=False)
+NonEmptyString = Annotated[
+    NonEmptyStringBase,
+    Field(description="A non-empty string", min_length=1),
+]
+non_empty_string_adapter = TypeAdapter(NonEmptyString)
+
+
+def parse_non_empty_string(s: str) -> NonEmptyString:
+    return non_empty_string_adapter.validate_python(s)
+
+
+def try_parse_non_empty_string(s: str) -> NonEmptyString | None:
+    if not s:
+        return None
+    return parse_non_empty_string(s)

server/reflector/views/rooms.py

@@ -15,6 +15,7 @@ from reflector.db.meetings import meetings_controller
 from reflector.db.rooms import rooms_controller
 from reflector.settings import settings
 from reflector.whereby import create_meeting, upload_logo
+from reflector.worker.webhook import test_webhook
 
 logger = logging.getLogger(__name__)
 
@@ -44,6 +45,11 @@ class Room(BaseModel):
     is_shared: bool
 
 
+class RoomDetails(Room):
+    webhook_url: str | None
+    webhook_secret: str | None
+
+
 class Meeting(BaseModel):
     id: str
     room_name: str
@@ -64,6 +70,8 @@ class CreateRoom(BaseModel):
     recording_type: str
     recording_trigger: str
     is_shared: bool
+    webhook_url: str
+    webhook_secret: str
 
 
 class UpdateRoom(BaseModel):
@@ -76,16 +84,26 @@ class UpdateRoom(BaseModel):
     recording_type: str
     recording_trigger: str
     is_shared: bool
+    webhook_url: str
+    webhook_secret: str
 
 
 class DeletionStatus(BaseModel):
     status: str
 
 
-@router.get("/rooms", response_model=Page[Room])
+class WebhookTestResult(BaseModel):
+    success: bool
+    message: str = ""
+    error: str = ""
+    status_code: int | None = None
+    response_preview: str | None = None
+
+
+@router.get("/rooms", response_model=Page[RoomDetails])
 async def rooms_list(
     user: Annotated[Optional[auth.UserInfo], Depends(auth.current_user_optional)],
-) -> list[Room]:
+) -> list[RoomDetails]:
     if not user and not settings.PUBLIC_MODE:
         raise HTTPException(status_code=401, detail="Not authenticated")
 
@@ -99,6 +117,18 @@ async def rooms_list(
     )
 
 
+@router.get("/rooms/{room_id}", response_model=RoomDetails)
+async def rooms_get(
+    room_id: str,
+    user: Annotated[Optional[auth.UserInfo], Depends(auth.current_user_optional)],
+):
+    user_id = user["sub"] if user else None
+    room = await rooms_controller.get_by_id_for_http(room_id, user_id=user_id)
+    if not room:
+        raise HTTPException(status_code=404, detail="Room not found")
+    return room
+
+
 @router.post("/rooms", response_model=Room)
 async def rooms_create(
     room: CreateRoom,
@@ -117,10 +147,12 @@ async def rooms_create(
         recording_type=room.recording_type,
         recording_trigger=room.recording_trigger,
         is_shared=room.is_shared,
+        webhook_url=room.webhook_url,
+        webhook_secret=room.webhook_secret,
     )
 
 
-@router.patch("/rooms/{room_id}", response_model=Room)
+@router.patch("/rooms/{room_id}", response_model=RoomDetails)
 async def rooms_update(
     room_id: str,
     info: UpdateRoom,
@@ -165,6 +197,7 @@ async def rooms_create_meeting(
         end_date = current_time + timedelta(hours=8)
 
         whereby_meeting = await create_meeting("", end_date=end_date, room=room)
+
         await upload_logo(whereby_meeting["roomName"], "./images/logo.png")
 
         # Now try to save to database
@@ -209,3 +242,24 @@ async def rooms_create_meeting(
         meeting.host_room_url = ""
 
     return meeting
+
+
+@router.post("/rooms/{room_id}/webhook/test", response_model=WebhookTestResult)
+async def rooms_test_webhook(
+    room_id: str,
+    user: Annotated[Optional[auth.UserInfo], Depends(auth.current_user_optional)],
+):
+    """Test webhook configuration by sending a sample payload."""
+    user_id = user["sub"] if user else None
+
+    room = await rooms_controller.get_by_id(room_id)
+    if not room:
+        raise HTTPException(status_code=404, detail="Room not found")
+
+    if user_id and room.user_id != user_id:
+        raise HTTPException(
+            status_code=403, detail="Not authorized to test this room's webhook"
+        )
+
+    result = await test_webhook(room_id)
+    return WebhookTestResult(**result)

server/reflector/views/transcripts.py

@@ -5,7 +5,7 @@ from fastapi import APIRouter, Depends, HTTPException, Query
 from fastapi_pagination import Page
 from fastapi_pagination.ext.databases import apaginate
 from jose import jwt
-from pydantic import BaseModel, Field, field_serializer
+from pydantic import BaseModel, Field, constr, field_serializer
 
 import reflector.auth as auth
 from reflector.db import get_database
@@ -19,14 +19,15 @@ from reflector.db.search import (
     SearchOffsetBase,
     SearchParameters,
     SearchQuery,
-    SearchQueryBase,
     SearchResult,
     SearchTotal,
     search_controller,
+    search_query_adapter,
 )
 from reflector.db.transcripts import (
     SourceKind,
     TranscriptParticipant,
+    TranscriptStatus,
     TranscriptTopic,
     transcripts_controller,
 )
@@ -63,7 +64,7 @@ class GetTranscriptMinimal(BaseModel):
     id: str
     user_id: str | None
     name: str
-    status: str
+    status: TranscriptStatus
     locked: bool
     duration: float
     title: str | None
@@ -96,6 +97,7 @@ class CreateTranscript(BaseModel):
     name: str
     source_language: str = Field("en")
     target_language: str = Field("en")
+    source_kind: SourceKind | None = None
 
 
 class UpdateTranscript(BaseModel):
@@ -114,7 +116,19 @@ class DeletionStatus(BaseModel):
     status: str
 
 
-SearchQueryParam = Annotated[SearchQueryBase, Query(description="Search query text")]
+SearchQueryParamBase = constr(min_length=0, strip_whitespace=True)
+SearchQueryParam = Annotated[
+    SearchQueryParamBase, Query(description="Search query text")
+]
+
+
+# http and api standards accept "q="; we would like to handle it as the absence of query, not as "empty string query"
+def parse_search_query_param(q: SearchQueryParam) -> SearchQuery | None:
+    if q == "":
+        return None
+    return search_query_adapter.validate_python(q)
+
+
 SearchLimitParam = Annotated[SearchLimitBase, Query(description="Results per page")]
 SearchOffsetParam = Annotated[
     SearchOffsetBase, Query(description="Number of results to skip")
@@ -124,7 +138,7 @@ SearchOffsetParam = Annotated[
 class SearchResponse(BaseModel):
     results: list[SearchResult]
     total: SearchTotal
-    query: SearchQuery
+    query: SearchQuery | None = None
     limit: SearchLimit
     offset: SearchOffset
 
@@ -174,7 +188,7 @@ async def transcripts_search(
     user_id = user["sub"] if user else None
 
     search_params = SearchParameters(
-        query_text=q,
+        query_text=parse_search_query_param(q),
         limit=limit,
         offset=offset,
         user_id=user_id,
@@ -201,7 +215,7 @@ async def transcripts_create(
     user_id = user["sub"] if user else None
     return await transcripts_controller.add(
         info.name,
-        source_kind=SourceKind.LIVE,
+        source_kind=info.source_kind or SourceKind.LIVE,
         source_language=info.source_language,
         target_language=info.target_language,
         user_id=user_id,

server/reflector/views/transcripts_process.py

@@ -6,7 +6,7 @@ from pydantic import BaseModel
 
 import reflector.auth as auth
 from reflector.db.transcripts import transcripts_controller
-from reflector.pipelines.main_live_pipeline import task_pipeline_process
+from reflector.pipelines.main_file_pipeline import task_pipeline_file_process
 
 router = APIRouter()
 
@@ -34,13 +34,13 @@ async def transcript_process(
         )
 
     if task_is_scheduled_or_active(
-        "reflector.pipelines.main_live_pipeline.task_pipeline_process",
+        "reflector.pipelines.main_file_pipeline.task_pipeline_file_process",
         transcript_id=transcript_id,
     ):
         return ProcessStatus(status="already running")
 
     # schedule a background task process the file
-    task_pipeline_process.delay(transcript_id=transcript_id)
+    task_pipeline_file_process.delay(transcript_id=transcript_id)
 
     return ProcessStatus(status="ok")
 

server/reflector/views/transcripts_upload.py

@@ -6,7 +6,7 @@ from pydantic import BaseModel
 
 import reflector.auth as auth
 from reflector.db.transcripts import transcripts_controller
-from reflector.pipelines.main_live_pipeline import task_pipeline_process
+from reflector.pipelines.main_file_pipeline import task_pipeline_file_process
 
 router = APIRouter()
 
@@ -92,6 +92,6 @@ async def transcript_record_upload(
     await transcripts_controller.update(transcript, {"status": "uploaded"})
 
     # launch a background task to process the file
-    task_pipeline_process.delay(transcript_id=transcript_id)
+    task_pipeline_file_process.delay(transcript_id=transcript_id)
 
     return UploadStatus(status="ok")

server/reflector/worker/app.py

@@ -19,6 +19,7 @@ else:
             "reflector.pipelines.main_live_pipeline",
             "reflector.worker.healthcheck",
             "reflector.worker.process",
+            "reflector.worker.cleanup",
         ]
     )
 
@@ -38,6 +39,16 @@ else:
         },
     }
 
+    if settings.PUBLIC_MODE:
+        app.conf.beat_schedule["cleanup_old_public_data"] = {
+            "task": "reflector.worker.cleanup.cleanup_old_public_data_task",
+            "schedule": crontab(hour=3, minute=0),
+        }
+        logger.info(
+            "Public mode cleanup enabled",
+            retention_days=settings.PUBLIC_DATA_RETENTION_DAYS,
+        )
+
     if settings.HEALTHCHECK_URL:
         app.conf.beat_schedule["healthcheck_ping"] = {
             "task": "reflector.worker.healthcheck.healthcheck_ping",

server/reflector/worker/cleanup.py

@@ -0,0 +1,156 @@
+"""
+Main task for cleanup old public data.
+
+Deletes old anonymous transcripts and their associated meetings/recordings.
+Transcripts are the main entry point - any associated data is also removed.
+"""
+
+import asyncio
+from datetime import datetime, timedelta, timezone
+from typing import TypedDict
+
+import structlog
+from celery import shared_task
+from databases import Database
+from pydantic.types import PositiveInt
+
+from reflector.asynctask import asynctask
+from reflector.db import get_database
+from reflector.db.meetings import meetings
+from reflector.db.recordings import recordings
+from reflector.db.transcripts import transcripts, transcripts_controller
+from reflector.settings import settings
+from reflector.storage import get_recordings_storage
+
+logger = structlog.get_logger(__name__)
+
+
+class CleanupStats(TypedDict):
+    """Statistics for cleanup operation."""
+
+    transcripts_deleted: int
+    meetings_deleted: int
+    recordings_deleted: int
+    errors: list[str]
+
+
+async def delete_single_transcript(
+    db: Database, transcript_data: dict, stats: CleanupStats
+):
+    transcript_id = transcript_data["id"]
+    meeting_id = transcript_data["meeting_id"]
+    recording_id = transcript_data["recording_id"]
+
+    try:
+        async with db.transaction(isolation="serializable"):
+            if meeting_id:
+                await db.execute(meetings.delete().where(meetings.c.id == meeting_id))
+                stats["meetings_deleted"] += 1
+                logger.info("Deleted associated meeting", meeting_id=meeting_id)
+
+            if recording_id:
+                recording = await db.fetch_one(
+                    recordings.select().where(recordings.c.id == recording_id)
+                )
+                if recording:
+                    try:
+                        await get_recordings_storage().delete_file(
+                            recording["object_key"]
+                        )
+                    except Exception as storage_error:
+                        logger.warning(
+                            "Failed to delete recording from storage",
+                            recording_id=recording_id,
+                            object_key=recording["object_key"],
+                            error=str(storage_error),
+                        )
+
+                    await db.execute(
+                        recordings.delete().where(recordings.c.id == recording_id)
+                    )
+                    stats["recordings_deleted"] += 1
+                    logger.info(
+                        "Deleted associated recording", recording_id=recording_id
+                    )
+
+            await transcripts_controller.remove_by_id(transcript_id)
+            stats["transcripts_deleted"] += 1
+            logger.info(
+                "Deleted transcript",
+                transcript_id=transcript_id,
+                created_at=transcript_data["created_at"].isoformat(),
+            )
+    except Exception as e:
+        error_msg = f"Failed to delete transcript {transcript_id}: {str(e)}"
+        logger.error(error_msg, exc_info=e)
+        stats["errors"].append(error_msg)
+
+
+async def cleanup_old_transcripts(
+    db: Database, cutoff_date: datetime, stats: CleanupStats
+):
+    """Delete old anonymous transcripts and their associated recordings/meetings."""
+    query = transcripts.select().where(
+        (transcripts.c.created_at < cutoff_date) & (transcripts.c.user_id.is_(None))
+    )
+    old_transcripts = await db.fetch_all(query)
+
+    logger.info(f"Found {len(old_transcripts)} old transcripts to delete")
+
+    for transcript_data in old_transcripts:
+        await delete_single_transcript(db, transcript_data, stats)
+
+
+def log_cleanup_results(stats: CleanupStats):
+    logger.info(
+        "Cleanup completed",
+        transcripts_deleted=stats["transcripts_deleted"],
+        meetings_deleted=stats["meetings_deleted"],
+        recordings_deleted=stats["recordings_deleted"],
+        errors_count=len(stats["errors"]),
+    )
+
+    if stats["errors"]:
+        logger.warning(
+            "Cleanup completed with errors",
+            errors=stats["errors"][:10],
+        )
+
+
+async def cleanup_old_public_data(
+    days: PositiveInt | None = None,
+) -> CleanupStats | None:
+    if days is None:
+        days = settings.PUBLIC_DATA_RETENTION_DAYS
+
+    if not settings.PUBLIC_MODE:
+        logger.info("Skipping cleanup - not a public instance")
+        return None
+
+    cutoff_date = datetime.now(timezone.utc) - timedelta(days=days)
+    logger.info(
+        "Starting cleanup of old public data",
+        cutoff_date=cutoff_date.isoformat(),
+    )
+
+    stats: CleanupStats = {
+        "transcripts_deleted": 0,
+        "meetings_deleted": 0,
+        "recordings_deleted": 0,
+        "errors": [],
+    }
+
+    db = get_database()
+    await cleanup_old_transcripts(db, cutoff_date, stats)
+
+    log_cleanup_results(stats)
+    return stats
+
+
+@shared_task(
+    autoretry_for=(Exception,),
+    retry_kwargs={"max_retries": 3, "countdown": 300},
+)
+@asynctask
+def cleanup_old_public_data_task(days: int | None = None):
+    asyncio.run(cleanup_old_public_data(days=days))

server/reflector/worker/webhook.py

@@ -0,0 +1,258 @@
+"""Webhook task for sending transcript notifications."""
+
+import hashlib
+import hmac
+import json
+import uuid
+from datetime import datetime, timezone
+
+import httpx
+import structlog
+from celery import shared_task
+from celery.utils.log import get_task_logger
+
+from reflector.db.rooms import rooms_controller
+from reflector.db.transcripts import transcripts_controller
+from reflector.pipelines.main_live_pipeline import asynctask
+from reflector.settings import settings
+from reflector.utils.webvtt import topics_to_webvtt
+
+logger = structlog.wrap_logger(get_task_logger(__name__))
+
+
+def generate_webhook_signature(payload: bytes, secret: str, timestamp: str) -> str:
+    """Generate HMAC signature for webhook payload."""
+    signed_payload = f"{timestamp}.{payload.decode('utf-8')}"
+    hmac_obj = hmac.new(
+        secret.encode("utf-8"),
+        signed_payload.encode("utf-8"),
+        hashlib.sha256,
+    )
+    return hmac_obj.hexdigest()
+
+
+@shared_task(
+    bind=True,
+    max_retries=30,
+    default_retry_delay=60,
+    retry_backoff=True,
+    retry_backoff_max=3600,  # Max 1 hour between retries
+)
+@asynctask
+async def send_transcript_webhook(
+    self,
+    transcript_id: str,
+    room_id: str,
+    event_id: str,
+):
+    log = logger.bind(
+        transcript_id=transcript_id,
+        room_id=room_id,
+        retry_count=self.request.retries,
+    )
+
+    try:
+        # Fetch transcript and room
+        transcript = await transcripts_controller.get_by_id(transcript_id)
+        if not transcript:
+            log.error("Transcript not found, skipping webhook")
+            return
+
+        room = await rooms_controller.get_by_id(room_id)
+        if not room:
+            log.error("Room not found, skipping webhook")
+            return
+
+        if not room.webhook_url:
+            log.info("No webhook URL configured for room, skipping")
+            return
+
+        # Generate WebVTT content from topics
+        topics_data = []
+
+        if transcript.topics:
+            # Build topics data with diarized content per topic
+            for topic in transcript.topics:
+                topic_webvtt = topics_to_webvtt([topic]) if topic.words else ""
+                topics_data.append(
+                    {
+                        "title": topic.title,
+                        "summary": topic.summary,
+                        "timestamp": topic.timestamp,
+                        "duration": topic.duration,
+                        "webvtt": topic_webvtt,
+                    }
+                )
+
+        # Build webhook payload
+        frontend_url = f"{settings.UI_BASE_URL}/transcripts/{transcript.id}"
+        participants = [
+            {"id": p.id, "name": p.name, "speaker": p.speaker}
+            for p in (transcript.participants or [])
+        ]
+        payload_data = {
+            "event": "transcript.completed",
+            "event_id": event_id,
+            "timestamp": datetime.now(timezone.utc).isoformat(),
+            "transcript": {
+                "id": transcript.id,
+                "room_id": transcript.room_id,
+                "created_at": transcript.created_at.isoformat(),
+                "duration": transcript.duration,
+                "title": transcript.title,
+                "short_summary": transcript.short_summary,
+                "long_summary": transcript.long_summary,
+                "webvtt": transcript.webvtt,
+                "topics": topics_data,
+                "participants": participants,
+                "source_language": transcript.source_language,
+                "target_language": transcript.target_language,
+                "status": transcript.status,
+                "frontend_url": frontend_url,
+            },
+            "room": {
+                "id": room.id,
+                "name": room.name,
+            },
+        }
+
+        # Convert to JSON
+        payload_json = json.dumps(payload_data, separators=(",", ":"))
+        payload_bytes = payload_json.encode("utf-8")
+
+        # Generate signature if secret is configured
+        headers = {
+            "Content-Type": "application/json",
+            "User-Agent": "Reflector-Webhook/1.0",
+            "X-Webhook-Event": "transcript.completed",
+            "X-Webhook-Retry": str(self.request.retries),
+        }
+
+        if room.webhook_secret:
+            timestamp = str(int(datetime.now(timezone.utc).timestamp()))
+            signature = generate_webhook_signature(
+                payload_bytes, room.webhook_secret, timestamp
+            )
+            headers["X-Webhook-Signature"] = f"t={timestamp},v1={signature}"
+
+        # Send webhook with timeout
+        async with httpx.AsyncClient(timeout=30.0) as client:
+            log.info(
+                "Sending webhook",
+                url=room.webhook_url,
+                payload_size=len(payload_bytes),
+            )
+
+            response = await client.post(
+                room.webhook_url,
+                content=payload_bytes,
+                headers=headers,
+            )
+
+            response.raise_for_status()
+
+            log.info(
+                "Webhook sent successfully",
+                status_code=response.status_code,
+                response_size=len(response.content),
+            )
+
+    except httpx.HTTPStatusError as e:
+        log.error(
+            "Webhook failed with HTTP error",
+            status_code=e.response.status_code,
+            response_text=e.response.text[:500],  # First 500 chars
+        )
+
+        # Don't retry on client errors (4xx)
+        if 400 <= e.response.status_code < 500:
+            log.error("Client error, not retrying")
+            return
+
+        # Retry on server errors (5xx)
+        raise self.retry(exc=e)
+
+    except (httpx.ConnectError, httpx.TimeoutException) as e:
+        # Retry on network errors
+        log.error("Webhook failed with connection error", error=str(e))
+        raise self.retry(exc=e)
+
+    except Exception as e:
+        # Retry on unexpected errors
+        log.exception("Unexpected error in webhook task", error=str(e))
+        raise self.retry(exc=e)
+
+
+async def test_webhook(room_id: str) -> dict:
+    """
+    Test webhook configuration by sending a sample payload.
+    Returns immediately with success/failure status.
+    This is the shared implementation used by both the API endpoint and Celery task.
+    """
+    try:
+        room = await rooms_controller.get_by_id(room_id)
+        if not room:
+            return {"success": False, "error": "Room not found"}
+
+        if not room.webhook_url:
+            return {"success": False, "error": "No webhook URL configured"}
+
+        now = (datetime.now(timezone.utc).isoformat(),)
+        payload_data = {
+            "event": "test",
+            "event_id": uuid.uuid4().hex,
+            "timestamp": now,
+            "message": "This is a test webhook from Reflector",
+            "room": {
+                "id": room.id,
+                "name": room.name,
+            },
+        }
+
+        payload_json = json.dumps(payload_data, separators=(",", ":"))
+        payload_bytes = payload_json.encode("utf-8")
+
+        # Generate headers with signature
+        headers = {
+            "Content-Type": "application/json",
+            "User-Agent": "Reflector-Webhook/1.0",
+            "X-Webhook-Event": "test",
+        }
+
+        if room.webhook_secret:
+            timestamp = str(int(datetime.now(timezone.utc).timestamp()))
+            signature = generate_webhook_signature(
+                payload_bytes, room.webhook_secret, timestamp
+            )
+            headers["X-Webhook-Signature"] = f"t={timestamp},v1={signature}"
+
+        # Send test webhook with short timeout
+        async with httpx.AsyncClient(timeout=10.0) as client:
+            response = await client.post(
+                room.webhook_url,
+                content=payload_bytes,
+                headers=headers,
+            )
+
+            return {
+                "success": response.is_success,
+                "status_code": response.status_code,
+                "message": f"Webhook test {'successful' if response.is_success else 'failed'}",
+                "response_preview": response.text if response.text else None,
+            }
+
+    except httpx.TimeoutException:
+        return {
+            "success": False,
+            "error": "Webhook request timed out (10 seconds)",
+        }
+    except httpx.ConnectError as e:
+        return {
+            "success": False,
+            "error": f"Could not connect to webhook URL: {str(e)}",
+        }
+    except Exception as e:
+        return {
+            "success": False,
+            "error": f"Unexpected error: {str(e)}",
+        }

server/runserver.sh

@@ -2,7 +2,7 @@
 
 if [ "${ENTRYPOINT}" = "server" ]; then
     uv run alembic upgrade head
-    uv run -m reflector.app
+    uv run uvicorn reflector.app:app --host 0.0.0.0 --port 1250
 elif [ "${ENTRYPOINT}" = "worker" ]; then
     uv run celery -A reflector.worker.app worker --loglevel=info
 elif [ "${ENTRYPOINT}" = "beat" ]; then

server/tests/conftest.py

@@ -178,6 +178,63 @@ async def dummy_diarization():
         yield
 
 
+@pytest.fixture
+async def dummy_file_transcript():
+    from reflector.processors.file_transcript import FileTranscriptProcessor
+    from reflector.processors.types import Transcript, Word
+
+    class TestFileTranscriptProcessor(FileTranscriptProcessor):
+        async def _transcript(self, data):
+            return Transcript(
+                text="Hello world. How are you today?",
+                words=[
+                    Word(start=0.0, end=0.5, text="Hello", speaker=0),
+                    Word(start=0.5, end=0.6, text=" ", speaker=0),
+                    Word(start=0.6, end=1.0, text="world", speaker=0),
+                    Word(start=1.0, end=1.1, text=".", speaker=0),
+                    Word(start=1.1, end=1.2, text=" ", speaker=0),
+                    Word(start=1.2, end=1.5, text="How", speaker=0),
+                    Word(start=1.5, end=1.6, text=" ", speaker=0),
+                    Word(start=1.6, end=1.8, text="are", speaker=0),
+                    Word(start=1.8, end=1.9, text=" ", speaker=0),
+                    Word(start=1.9, end=2.1, text="you", speaker=0),
+                    Word(start=2.1, end=2.2, text=" ", speaker=0),
+                    Word(start=2.2, end=2.5, text="today", speaker=0),
+                    Word(start=2.5, end=2.6, text="?", speaker=0),
+                ],
+            )
+
+    with patch(
+        "reflector.processors.file_transcript_auto.FileTranscriptAutoProcessor.__new__"
+    ) as mock_auto:
+        mock_auto.return_value = TestFileTranscriptProcessor()
+        yield
+
+
+@pytest.fixture
+async def dummy_file_diarization():
+    from reflector.processors.file_diarization import (
+        FileDiarizationOutput,
+        FileDiarizationProcessor,
+    )
+    from reflector.processors.types import DiarizationSegment
+
+    class TestFileDiarizationProcessor(FileDiarizationProcessor):
+        async def _diarize(self, data):
+            return FileDiarizationOutput(
+                diarization=[
+                    DiarizationSegment(start=0.0, end=1.1, speaker=0),
+                    DiarizationSegment(start=1.2, end=2.6, speaker=1),
+                ]
+            )
+
+    with patch(
+        "reflector.processors.file_diarization_auto.FileDiarizationAutoProcessor.__new__"
+    ) as mock_auto:
+        mock_auto.return_value = TestFileDiarizationProcessor()
+        yield
+
+
 @pytest.fixture
 async def dummy_transcript_translator():
     from reflector.processors.transcript_translator import TranscriptTranslatorProcessor
@@ -238,9 +295,13 @@ async def dummy_storage():
     with (
         patch("reflector.storage.base.Storage.get_instance") as mock_storage,
         patch("reflector.storage.get_transcripts_storage") as mock_get_transcripts,
+        patch(
+            "reflector.pipelines.main_file_pipeline.get_transcripts_storage"
+        ) as mock_get_transcripts2,
     ):
         mock_storage.return_value = dummy
         mock_get_transcripts.return_value = dummy
+        mock_get_transcripts2.return_value = dummy
         yield
 
 
@@ -260,7 +321,10 @@ def celery_config():
 
 @pytest.fixture(scope="session")
 def celery_includes():
-    return ["reflector.pipelines.main_live_pipeline"]
+    return [
+        "reflector.pipelines.main_live_pipeline",
+        "reflector.pipelines.main_file_pipeline",
+    ]
 
 
 @pytest.fixture
@@ -302,7 +366,7 @@ async def fake_transcript_with_topics(tmpdir, client):
     transcript = await transcripts_controller.get_by_id(tid)
     assert transcript is not None
 
-    await transcripts_controller.update(transcript, {"status": "finished"})
+    await transcripts_controller.update(transcript, {"status": "ended"})
 
     # manually copy a file at the expected location
     audio_filename = transcript.audio_mp3_filename

server/tests/test_cleanup.py

@@ -0,0 +1,287 @@
+from datetime import datetime, timedelta, timezone
+from unittest.mock import AsyncMock, patch
+
+import pytest
+
+from reflector.db.recordings import Recording, recordings_controller
+from reflector.db.transcripts import SourceKind, transcripts_controller
+from reflector.worker.cleanup import cleanup_old_public_data
+
+
+@pytest.mark.asyncio
+async def test_cleanup_old_public_data_skips_when_not_public():
+    """Test that cleanup is skipped when PUBLIC_MODE is False."""
+    with patch("reflector.worker.cleanup.settings") as mock_settings:
+        mock_settings.PUBLIC_MODE = False
+
+        result = await cleanup_old_public_data()
+
+        # Should return early without doing anything
+        assert result is None
+
+
+@pytest.mark.asyncio
+async def test_cleanup_old_public_data_deletes_old_anonymous_transcripts():
+    """Test that old anonymous transcripts are deleted."""
+    # Create old and new anonymous transcripts
+    old_date = datetime.now(timezone.utc) - timedelta(days=8)
+    new_date = datetime.now(timezone.utc) - timedelta(days=2)
+
+    # Create old anonymous transcript (should be deleted)
+    old_transcript = await transcripts_controller.add(
+        name="Old Anonymous Transcript",
+        source_kind=SourceKind.FILE,
+        user_id=None,  # Anonymous
+    )
+    # Manually update created_at to be old
+    from reflector.db import get_database
+    from reflector.db.transcripts import transcripts
+
+    await get_database().execute(
+        transcripts.update()
+        .where(transcripts.c.id == old_transcript.id)
+        .values(created_at=old_date)
+    )
+
+    # Create new anonymous transcript (should NOT be deleted)
+    new_transcript = await transcripts_controller.add(
+        name="New Anonymous Transcript",
+        source_kind=SourceKind.FILE,
+        user_id=None,  # Anonymous
+    )
+
+    # Create old transcript with user (should NOT be deleted)
+    old_user_transcript = await transcripts_controller.add(
+        name="Old User Transcript",
+        source_kind=SourceKind.FILE,
+        user_id="user123",
+    )
+    await get_database().execute(
+        transcripts.update()
+        .where(transcripts.c.id == old_user_transcript.id)
+        .values(created_at=old_date)
+    )
+
+    with patch("reflector.worker.cleanup.settings") as mock_settings:
+        mock_settings.PUBLIC_MODE = True
+        mock_settings.PUBLIC_DATA_RETENTION_DAYS = 7
+
+        # Mock the storage deletion
+        with patch("reflector.db.transcripts.get_transcripts_storage") as mock_storage:
+            mock_storage.return_value.delete_file = AsyncMock()
+
+            result = await cleanup_old_public_data()
+
+    # Check results
+    assert result["transcripts_deleted"] == 1
+    assert result["errors"] == []
+
+    # Verify old anonymous transcript was deleted
+    assert await transcripts_controller.get_by_id(old_transcript.id) is None
+
+    # Verify new anonymous transcript still exists
+    assert await transcripts_controller.get_by_id(new_transcript.id) is not None
+
+    # Verify user transcript still exists
+    assert await transcripts_controller.get_by_id(old_user_transcript.id) is not None
+
+
+@pytest.mark.asyncio
+async def test_cleanup_deletes_associated_meeting_and_recording():
+    """Test that meetings and recordings associated with old transcripts are deleted."""
+    from reflector.db import get_database
+    from reflector.db.meetings import meetings
+    from reflector.db.transcripts import transcripts
+
+    old_date = datetime.now(timezone.utc) - timedelta(days=8)
+
+    # Create a meeting
+    meeting_id = "test-meeting-for-transcript"
+    await get_database().execute(
+        meetings.insert().values(
+            id=meeting_id,
+            room_name="Meeting with Transcript",
+            room_url="https://example.com/meeting",
+            host_room_url="https://example.com/meeting-host",
+            start_date=old_date,
+            end_date=old_date + timedelta(hours=1),
+            user_id=None,
+            room_id=None,
+        )
+    )
+
+    # Create a recording
+    recording = await recordings_controller.create(
+        Recording(
+            bucket_name="test-bucket",
+            object_key="test-recording.mp4",
+            recorded_at=old_date,
+        )
+    )
+
+    # Create an old transcript with both meeting and recording
+    old_transcript = await transcripts_controller.add(
+        name="Old Transcript with Meeting and Recording",
+        source_kind=SourceKind.ROOM,
+        user_id=None,
+        meeting_id=meeting_id,
+        recording_id=recording.id,
+    )
+
+    # Update created_at to be old
+    await get_database().execute(
+        transcripts.update()
+        .where(transcripts.c.id == old_transcript.id)
+        .values(created_at=old_date)
+    )
+
+    with patch("reflector.worker.cleanup.settings") as mock_settings:
+        mock_settings.PUBLIC_MODE = True
+        mock_settings.PUBLIC_DATA_RETENTION_DAYS = 7
+
+        # Mock storage deletion
+        with patch("reflector.db.transcripts.get_transcripts_storage") as mock_storage:
+            mock_storage.return_value.delete_file = AsyncMock()
+            with patch(
+                "reflector.worker.cleanup.get_recordings_storage"
+            ) as mock_rec_storage:
+                mock_rec_storage.return_value.delete_file = AsyncMock()
+
+                result = await cleanup_old_public_data()
+
+    # Check results
+    assert result["transcripts_deleted"] == 1
+    assert result["meetings_deleted"] == 1
+    assert result["recordings_deleted"] == 1
+    assert result["errors"] == []
+
+    # Verify transcript was deleted
+    assert await transcripts_controller.get_by_id(old_transcript.id) is None
+
+    # Verify meeting was deleted
+    query = meetings.select().where(meetings.c.id == meeting_id)
+    meeting_result = await get_database().fetch_one(query)
+    assert meeting_result is None
+
+    # Verify recording was deleted
+    assert await recordings_controller.get_by_id(recording.id) is None
+
+
+@pytest.mark.asyncio
+async def test_cleanup_handles_errors_gracefully():
+    """Test that cleanup continues even when individual deletions fail."""
+    old_date = datetime.now(timezone.utc) - timedelta(days=8)
+
+    # Create multiple old transcripts
+    transcript1 = await transcripts_controller.add(
+        name="Transcript 1",
+        source_kind=SourceKind.FILE,
+        user_id=None,
+    )
+    transcript2 = await transcripts_controller.add(
+        name="Transcript 2",
+        source_kind=SourceKind.FILE,
+        user_id=None,
+    )
+
+    # Update created_at to be old
+    from reflector.db import get_database
+    from reflector.db.transcripts import transcripts
+
+    for t_id in [transcript1.id, transcript2.id]:
+        await get_database().execute(
+            transcripts.update()
+            .where(transcripts.c.id == t_id)
+            .values(created_at=old_date)
+        )
+
+    with patch("reflector.worker.cleanup.settings") as mock_settings:
+        mock_settings.PUBLIC_MODE = True
+        mock_settings.PUBLIC_DATA_RETENTION_DAYS = 7
+
+        # Mock remove_by_id to fail for the first transcript
+        original_remove = transcripts_controller.remove_by_id
+        call_count = 0
+
+        async def mock_remove_by_id(transcript_id, user_id=None):
+            nonlocal call_count
+            call_count += 1
+            if call_count == 1:
+                raise Exception("Simulated deletion error")
+            return await original_remove(transcript_id, user_id)
+
+        with patch.object(
+            transcripts_controller, "remove_by_id", side_effect=mock_remove_by_id
+        ):
+            result = await cleanup_old_public_data()
+
+    # Should have one successful deletion and one error
+    assert result["transcripts_deleted"] == 1
+    assert len(result["errors"]) == 1
+    assert "Failed to delete transcript" in result["errors"][0]
+
+
+@pytest.mark.asyncio
+async def test_meeting_consent_cascade_delete():
+    """Test that meeting_consent records are automatically deleted when meeting is deleted."""
+    from reflector.db import get_database
+    from reflector.db.meetings import (
+        meeting_consent,
+        meeting_consent_controller,
+        meetings,
+    )
+
+    # Create a meeting
+    meeting_id = "test-cascade-meeting"
+    await get_database().execute(
+        meetings.insert().values(
+            id=meeting_id,
+            room_name="Test Meeting for CASCADE",
+            room_url="https://example.com/cascade-test",
+            host_room_url="https://example.com/cascade-test-host",
+            start_date=datetime.now(timezone.utc),
+            end_date=datetime.now(timezone.utc) + timedelta(hours=1),
+            user_id="test-user",
+            room_id=None,
+        )
+    )
+
+    # Create consent records for this meeting
+    consent1_id = "consent-1"
+    consent2_id = "consent-2"
+
+    await get_database().execute(
+        meeting_consent.insert().values(
+            id=consent1_id,
+            meeting_id=meeting_id,
+            user_id="user1",
+            consent_given=True,
+            consent_timestamp=datetime.now(timezone.utc),
+        )
+    )
+
+    await get_database().execute(
+        meeting_consent.insert().values(
+            id=consent2_id,
+            meeting_id=meeting_id,
+            user_id="user2",
+            consent_given=False,
+            consent_timestamp=datetime.now(timezone.utc),
+        )
+    )
+
+    # Verify consent records exist
+    consents = await meeting_consent_controller.get_by_meeting_id(meeting_id)
+    assert len(consents) == 2
+
+    # Delete the meeting
+    await get_database().execute(meetings.delete().where(meetings.c.id == meeting_id))
+
+    # Verify meeting is deleted
+    query = meetings.select().where(meetings.c.id == meeting_id)
+    result = await get_database().fetch_one(query)
+    assert result is None
+
+    # Verify consent records are automatically deleted (CASCADE DELETE)
+    consents_after = await meeting_consent_controller.get_by_meeting_id(meeting_id)
+    assert len(consents_after) == 0

server/tests/test_gpu_modal_transcript.py

@@ -272,6 +272,9 @@ class TestGPUModalTranscript:
                 for f in temp_files:
                     Path(f).unlink(missing_ok=True)
 
+    @pytest.mark.skipif(
+        not "parakeet" in get_model_name(), reason="Parakeet only supports English"
+    )
     def test_transcriptions_error_handling(self):
         """Test error handling for invalid requests."""
         url = get_modal_transcript_url()

server/tests/test_processors_pipeline.py

@@ -1,61 +0,0 @@
-import pytest
-
-
-@pytest.mark.asyncio
-@pytest.mark.parametrize("enable_diarization", [False, True])
-async def test_basic_process(
-    dummy_transcript,
-    dummy_llm,
-    dummy_processors,
-    enable_diarization,
-    dummy_diarization,
-):
-    # goal is to start the server, and send rtc audio to it
-    # validate the events received
-    from pathlib import Path
-
-    from reflector.settings import settings
-    from reflector.tools.process import process_audio_file
-
-    # LLM_BACKEND no longer exists in settings
-    # settings.LLM_BACKEND = "test"
-    settings.TRANSCRIPT_BACKEND = "whisper"
-
-    # event callback
-    marks = {}
-
-    async def event_callback(event):
-        if event.processor not in marks:
-            marks[event.processor] = 0
-        marks[event.processor] += 1
-
-    # invoke the process and capture events
-    path = Path(__file__).parent / "records" / "test_mathieu_hello.wav"
-
-    if enable_diarization:
-        # Test with diarization - may fail if pyannote.audio is not installed
-        try:
-            await process_audio_file(
-                path.as_posix(), event_callback, enable_diarization=True
-            )
-        except SystemExit:
-            pytest.skip("pyannote.audio not installed - skipping diarization test")
-    else:
-        # Test without diarization - should always work
-        await process_audio_file(
-            path.as_posix(), event_callback, enable_diarization=False
-        )
-
-    print(f"Diarization: {enable_diarization}, Marks: {marks}")
-
-    # validate the events
-    # Each processor should be called for each audio segment processed
-    # The final processors (Topic, Title, Summary) should be called once at the end
-    assert marks["TranscriptLinerProcessor"] > 0
-    assert marks["TranscriptTranslatorPassthroughProcessor"] > 0
-    assert marks["TranscriptTopicDetectorProcessor"] == 1
-    assert marks["TranscriptFinalSummaryProcessor"] == 1
-    assert marks["TranscriptFinalTitleProcessor"] == 1
-
-    if enable_diarization:
-        assert marks["TestAudioDiarizationProcessor"] == 1

server/tests/test_search.py

@@ -23,7 +23,7 @@ async def test_search_postgresql_only():
     assert results == []
     assert total == 0
 
-    params_empty = SearchParameters(query_text="")
+    params_empty = SearchParameters(query_text=None)
     results_empty, total_empty = await search_controller.search_transcripts(
         params_empty
     )
@@ -34,7 +34,7 @@ async def test_search_postgresql_only():
 @pytest.mark.asyncio
 async def test_search_with_empty_query():
     """Test that empty query returns all transcripts."""
-    params = SearchParameters(query_text="")
+    params = SearchParameters(query_text=None)
     results, total = await search_controller.search_transcripts(params)
 
     assert isinstance(results, list)

server/tests/test_search_snippets.py

@@ -1,5 +1,7 @@
 """Unit tests for search snippet generation."""
 
+import pytest
+
 from reflector.db.search import (
     SnippetCandidate,
     SnippetGenerator,
@@ -512,11 +514,9 @@ data visualization and data storage"""
         )
         assert webvtt_count == 3
 
-        snippets_empty, count_empty = SnippetGenerator.combine_sources(
-            None, None, "data", max_total=3
-        )
-        assert snippets_empty == []
-        assert count_empty == 0
+        # combine_sources requires at least one source to be present
+        with pytest.raises(AssertionError, match="At least one source must be present"):
+            SnippetGenerator.combine_sources(None, None, "data", max_total=3)
 
     def test_edge_cases(self):
         """Test edge cases for the pure functions."""

server/tests/test_transcripts_audio_download.py

@@ -19,7 +19,7 @@ async def fake_transcript(tmpdir, client):
     transcript = await transcripts_controller.get_by_id(tid)
     assert transcript is not None
 
-    await transcripts_controller.update(transcript, {"status": "finished"})
+    await transcripts_controller.update(transcript, {"status": "ended"})
 
     # manually copy a file at the expected location
     audio_filename = transcript.audio_mp3_filename

server/tests/test_transcripts_process.py

@@ -29,10 +29,10 @@ async def client(app_lifespan):
 @pytest.mark.asyncio
 async def test_transcript_process(
     tmpdir,
-    whisper_transcript,
     dummy_llm,
     dummy_processors,
-    dummy_diarization,
+    dummy_file_transcript,
+    dummy_file_diarization,
     dummy_storage,
     client,
 ):
@@ -56,8 +56,8 @@ async def test_transcript_process(
     assert response.status_code == 200
     assert response.json()["status"] == "ok"
 
-    # wait for processing to finish (max 10 minutes)
-    timeout_seconds = 600  # 10 minutes
+    # wait for processing to finish (max 1 minute)
+    timeout_seconds = 60
     start_time = time.monotonic()
     while (time.monotonic() - start_time) < timeout_seconds:
         # fetch the transcript and check if it is ended
@@ -75,9 +75,10 @@ async def test_transcript_process(
     )
     assert response.status_code == 200
     assert response.json()["status"] == "ok"
+    await asyncio.sleep(2)
 
-    # wait for processing to finish (max 10 minutes)
-    timeout_seconds = 600  # 10 minutes
+    # wait for processing to finish (max 1 minute)
+    timeout_seconds = 60
     start_time = time.monotonic()
     while (time.monotonic() - start_time) < timeout_seconds:
         # fetch the transcript and check if it is ended
@@ -99,4 +100,4 @@ async def test_transcript_process(
     response = await client.get(f"/transcripts/{tid}/topics")
     assert response.status_code == 200
     assert len(response.json()) == 1
-    assert "want to share" in response.json()[0]["transcript"]
+    assert "Hello world. How are you today?" in response.json()[0]["transcript"]

server/tests/test_transcripts_upload.py

@@ -12,7 +12,8 @@ async def test_transcript_upload_file(
     tmpdir,
     dummy_llm,
     dummy_processors,
-    dummy_diarization,
+    dummy_file_transcript,
+    dummy_file_diarization,
     dummy_storage,
     client,
 ):
@@ -36,8 +37,8 @@ async def test_transcript_upload_file(
     assert response.status_code == 200
     assert response.json()["status"] == "ok"
 
-    # wait the processing to finish (max 10 minutes)
-    timeout_seconds = 600  # 10 minutes
+    # wait the processing to finish (max 1 minute)
+    timeout_seconds = 60
     start_time = time.monotonic()
     while (time.monotonic() - start_time) < timeout_seconds:
         # fetch the transcript and check if it is ended
@@ -47,7 +48,7 @@ async def test_transcript_upload_file(
             break
         await asyncio.sleep(1)
     else:
-        pytest.fail(f"Processing timed out after {timeout_seconds} seconds")
+        return pytest.fail(f"Processing timed out after {timeout_seconds} seconds")
 
     # check the transcript is ended
     transcript = resp.json()
@@ -59,4 +60,4 @@ async def test_transcript_upload_file(
     response = await client.get(f"/transcripts/{tid}/topics")
     assert response.status_code == 200
     assert len(response.json()) == 1
-    assert "want to share" in response.json()[0]["transcript"]
+    assert "Hello world. How are you today?" in response.json()[0]["transcript"]

www/app/(app)/AuthWrapper.tsx

@@ -0,0 +1,30 @@
+"use client";
+
+import { Flex, Spinner } from "@chakra-ui/react";
+import { useAuth } from "../lib/AuthProvider";
+import { useLoginRequiredPages } from "../lib/useLoginRequiredPages";
+
+export default function AuthWrapper({
+  children,
+}: {
+  children: React.ReactNode;
+}) {
+  const auth = useAuth();
+  const redirectPath = useLoginRequiredPages();
+  const redirectHappens = !!redirectPath;
+
+  if (auth.status === "loading" || redirectHappens) {
+    return (
+      <Flex
+        flexDir="column"
+        alignItems="center"
+        justifyContent="center"
+        h="calc(100vh - 80px)" // Account for header height
+      >
+        <Spinner size="xl" color="blue.500" />
+      </Flex>
+    );
+  }
+
+  return <>{children}</>;
+}

www/app/(app)/browse/_components/FilterSidebar.tsx

@@ -1,7 +1,10 @@
 import React from "react";
 import { Box, Stack, Link, Heading } from "@chakra-ui/react";
 import NextLink from "next/link";
-import { Room, SourceKind } from "../../../api";
+import type { components } from "../../../reflector-api";
+
+type Room = components["schemas"]["Room"];
+type SourceKind = components["schemas"]["SourceKind"];
 
 interface FilterSidebarProps {
   rooms: Room[];
@@ -72,7 +75,7 @@ export default function FilterSidebar({
                 key={room.id}
                 as={NextLink}
                 href="#"
-                onClick={() => onFilterChange("room", room.id)}
+                onClick={() => onFilterChange("room" as SourceKind, room.id)}
                 color={
                   selectedSourceKind === "room" && selectedRoomId === room.id
                     ? "blue.500"

www/app/(app)/browse/_components/TranscriptCards.tsx

@@ -18,7 +18,10 @@ import {
   highlightMatches,
   generateTextFragment,
 } from "../../../lib/textHighlight";
-import { SearchResult } from "../../../api";
+import type { components } from "../../../reflector-api";
+
+type SearchResult = components["schemas"]["SearchResult"];
+type SourceKind = components["schemas"]["SourceKind"];
 
 interface TranscriptCardsProps {
   results: SearchResult[];
@@ -120,7 +123,7 @@ function TranscriptCard({
     : "N/A";
   const formattedDate = formatLocalDate(result.created_at);
   const source =
-    result.source_kind === "room"
+    result.source_kind === ("room" as SourceKind)
       ? result.room_name || result.room_id
       : result.source_kind;
 

www/app/(app)/browse/page.tsx

@@ -19,37 +19,33 @@ import {
   parseAsStringLiteral,
 } from "nuqs";
 import { LuX } from "react-icons/lu";
-import { useSearchTranscripts } from "../transcripts/useSearchTranscripts";
-import useSessionUser from "../../lib/useSessionUser";
-import { Room, SourceKind, SearchResult, $SourceKind } from "../../api";
-import useApi from "../../lib/useApi";
-import { useError } from "../../(errors)/errorContext";
+import type { components } from "../../reflector-api";
+
+type Room = components["schemas"]["Room"];
+type SourceKind = components["schemas"]["SourceKind"];
+type SearchResult = components["schemas"]["SearchResult"];
+import {
+  useRoomsList,
+  useTranscriptsSearch,
+  useTranscriptDelete,
+  useTranscriptProcess,
+} from "../../lib/apiHooks";
 import FilterSidebar from "./_components/FilterSidebar";
 import Pagination, {
   FIRST_PAGE,
   PaginationPage,
   parsePaginationPage,
   totalPages as getTotalPages,
+  paginationPageTo0Based,
 } from "./_components/Pagination";
 import TranscriptCards from "./_components/TranscriptCards";
 import DeleteTranscriptDialog from "./_components/DeleteTranscriptDialog";
 import { formatLocalDate } from "../../lib/time";
 import { RECORD_A_MEETING_URL } from "../../api/urls";
+import { useUserName } from "../../lib/useUserName";
 
 const SEARCH_FORM_QUERY_INPUT_NAME = "query" as const;
 
-const usePrefetchRooms = (setRooms: (rooms: Room[]) => void): void => {
-  const { setError } = useError();
-  const api = useApi();
-  useEffect(() => {
-    if (!api) return;
-    api
-      .v1RoomsList({ page: 1 })
-      .then((rooms) => setRooms(rooms.items))
-      .catch((err) => setError(err, "There was an error fetching the rooms"));
-  }, [api, setError]);
-};
-
 const SearchForm: React.FC<{
   setPage: (page: PaginationPage) => void;
   sourceKind: SourceKind | null;
@@ -69,7 +65,6 @@ const SearchForm: React.FC<{
   searchQuery,
   setSearchQuery,
 }) => {
-  // to keep the search input controllable + more fine grained control (urlSearchQuery is updated on submits)
   const [searchInputValue, setSearchInputValue] = useState(searchQuery || "");
   const handleSearchQuerySubmit = async (d: FormData) => {
     await setSearchQuery((d.get(SEARCH_FORM_QUERY_INPUT_NAME) as string) || "");
@@ -163,7 +158,6 @@ const UnderSearchFormFilterIndicators: React.FC<{
                 p="1px"
                 onClick={() => {
                   setSourceKind(null);
-                  // TODO questionable
                   setRoomId(null);
                 }}
                 _hover={{ bg: "blue.200" }}
@@ -209,7 +203,11 @@ export default function TranscriptBrowser() {
 
   const [urlSourceKind, setUrlSourceKind] = useQueryState(
     "source",
-    parseAsStringLiteral($SourceKind.enum).withOptions({
+    parseAsStringLiteral([
+      "room",
+      "live",
+      "file",
+    ] as const satisfies SourceKind[]).withOptions({
       shallow: false,
     }),
   );
@@ -229,46 +227,40 @@ export default function TranscriptBrowser() {
   useEffect(() => {
     const maybePage = parsePaginationPage(urlPage);
     if ("error" in maybePage) {
-      setPage(FIRST_PAGE).then(() => {
-        /*may be called n times we dont care*/
-      });
+      setPage(FIRST_PAGE).then(() => {});
       return;
     }
     _setSafePage(maybePage.value);
   }, [urlPage]);
 
-  const [rooms, setRooms] = useState<Room[]>([]);
-
   const pageSize = 20;
+
   const {
-    results,
-    totalCount: totalResults,
-    isLoading,
-    reload,
-  } = useSearchTranscripts(
-    urlSearchQuery,
-    {
-      roomIds: urlRoomId ? [urlRoomId] : null,
-      sourceKind: urlSourceKind,
-    },
-    {
-      pageSize,
-      page,
-    },
-  );
+    data: searchData,
+    isLoading: searchLoading,
+    refetch: reloadSearch,
+  } = useTranscriptsSearch(urlSearchQuery, {
+    limit: pageSize,
+    offset: paginationPageTo0Based(page) * pageSize,
+    room_id: urlRoomId || undefined,
+    source_kind: urlSourceKind || undefined,
+  });
+
+  const results = searchData?.results || [];
+  const totalResults = searchData?.total || 0;
+
+  // Fetch rooms
+  const { data: roomsData } = useRoomsList(1);
+  const rooms = roomsData?.items || [];
 
   const totalPages = getTotalPages(totalResults, pageSize);
 
-  const userName = useSessionUser().name;
+  const userName = useUserName();
   const [deletionLoading, setDeletionLoading] = useState(false);
-  const api = useApi();
-  const { setError } = useError();
   const cancelRef = React.useRef(null);
   const [transcriptToDeleteId, setTranscriptToDeleteId] =
     React.useState<string>();
 
-  usePrefetchRooms(setRooms);
-
   const handleFilterTranscripts = (
     sourceKind: SourceKind | null,
     roomId: string,
@@ -280,44 +272,37 @@ export default function TranscriptBrowser() {
 
   const onCloseDeletion = () => setTranscriptToDeleteId(undefined);
 
+  const deleteTranscript = useTranscriptDelete();
+  const processTranscript = useTranscriptProcess();
+
   const confirmDeleteTranscript = (transcriptId: string) => {
-    if (!api || deletionLoading) return;
+    if (deletionLoading) return;
     setDeletionLoading(true);
-    api
-      .v1TranscriptDelete({ transcriptId })
-      .then(() => {
-        setDeletionLoading(false);
-        onCloseDeletion();
-        reload();
-      })
-      .catch((err) => {
-        setDeletionLoading(false);
-        setError(err, "There was an error deleting the transcript");
-      });
+    deleteTranscript.mutate(
+      {
+        params: {
+          path: { transcript_id: transcriptId },
+        },
+      },
+      {
+        onSuccess: () => {
+          setDeletionLoading(false);
+          onCloseDeletion();
+          reloadSearch();
+        },
+        onError: () => {
+          setDeletionLoading(false);
+        },
+      },
+    );
   };
 
   const handleProcessTranscript = (transcriptId: string) => {
-    if (!api) {
-      console.error("API not available on handleProcessTranscript");
-      return;
-    }
-    api
-      .v1TranscriptProcess({ transcriptId })
-      .then((result) => {
-        const status =
-          result && typeof result === "object" && "status" in result
-            ? (result as { status: string }).status
-            : undefined;
-        if (status === "already running") {
-          setError(
-            new Error("Processing is already running, please wait"),
-            "Processing is already running, please wait",
-          );
-        }
-      })
-      .catch((err) => {
-        setError(err, "There was an error processing the transcript");
-      });
+    processTranscript.mutate({
+      params: {
+        path: { transcript_id: transcriptId },
+      },
+    });
   };
 
   const transcriptToDelete = results?.find(
@@ -332,7 +317,7 @@ export default function TranscriptBrowser() {
       ? transcriptToDelete.room_name || transcriptToDelete.room_id
       : transcriptToDelete?.source_kind;
 
-  if (isLoading && results.length === 0) {
+  if (searchLoading && results.length === 0) {
     return (
       <Flex
         flexDir="column"
@@ -360,7 +345,7 @@ export default function TranscriptBrowser() {
       >
         <Heading size="lg">
           {userName ? `${userName}'s Transcriptions` : "Your Transcriptions"}{" "}
-          {(isLoading || deletionLoading) && <Spinner size="sm" />}
+          {(searchLoading || deletionLoading) && <Spinner size="sm" />}
         </Heading>
       </Flex>
 
@@ -403,12 +388,12 @@ export default function TranscriptBrowser() {
           <TranscriptCards
             results={results}
             query={urlSearchQuery}
-            isLoading={isLoading}
+            isLoading={searchLoading}
             onDelete={setTranscriptToDeleteId}
             onReprocess={handleProcessTranscript}
           />
 
-          {!isLoading && results.length === 0 && (
+          {!searchLoading && results.length === 0 && (
             <EmptyResult searchQuery={urlSearchQuery} />
           )}
         </Flex>

www/app/(app)/layout.tsx

@@ -2,9 +2,8 @@ import { Container, Flex, Link } from "@chakra-ui/react";
 import { getConfig } from "../lib/edgeConfig";
 import NextLink from "next/link";
 import Image from "next/image";
-import About from "../(aboutAndPrivacy)/about";
-import Privacy from "../(aboutAndPrivacy)/privacy";
 import UserInfo from "../(auth)/userInfo";
+import AuthWrapper from "./AuthWrapper";
 import { RECORD_A_MEETING_URL } from "../api/urls";
 
 export default async function AppLayout({
@@ -90,7 +89,7 @@ export default async function AppLayout({
         </div>
       </Flex>
 
-      {children}
+      <AuthWrapper>{children}</AuthWrapper>
     </Container>
   );
 }

www/app/(app)/rooms/_components/RoomCards.tsx

@@ -12,7 +12,9 @@ import {
   HStack,
 } from "@chakra-ui/react";
 import { LuLink } from "react-icons/lu";
-import { Room } from "../../../api";
+import type { components } from "../../../reflector-api";
+
+type Room = components["schemas"]["Room"];
 import { RoomActionsMenu } from "./RoomActionsMenu";
 
 interface RoomCardsProps {

www/app/(app)/rooms/_components/RoomList.tsx

@@ -1,5 +1,7 @@
 import { Box, Heading, Text, VStack } from "@chakra-ui/react";
-import { Room } from "../../../api";
+import type { components } from "../../../reflector-api";
+
+type Room = components["schemas"]["Room"];
 import { RoomTable } from "./RoomTable";
 import { RoomCards } from "./RoomCards";
 

www/app/(app)/rooms/_components/RoomTable.tsx

@@ -9,7 +9,9 @@ import {
   Spinner,
 } from "@chakra-ui/react";
 import { LuLink } from "react-icons/lu";
-import { Room } from "../../../api";
+import type { components } from "../../../reflector-api";
+
+type Room = components["schemas"]["Room"];
 import { RoomActionsMenu } from "./RoomActionsMenu";
 
 interface RoomTableProps {

www/app/(app)/rooms/page.tsx

@@ -11,15 +11,28 @@ import {
   Input,
   Select,
   Spinner,
+  IconButton,
   createListCollection,
   useDisclosure,
 } from "@chakra-ui/react";
-import { useEffect, useState } from "react";
-import useApi from "../../lib/useApi";
+import { useEffect, useMemo, useState } from "react";
+import { LuEye, LuEyeOff } from "react-icons/lu";
 import useRoomList from "./useRoomList";
-import { ApiError, Room } from "../../api";
+import type { components } from "../../reflector-api";
+import {
+  useRoomCreate,
+  useRoomUpdate,
+  useRoomDelete,
+  useZulipStreams,
+  useZulipTopics,
+  useRoomGet,
+  useRoomTestWebhook,
+} from "../../lib/apiHooks";
 import { RoomList } from "./_components/RoomList";
 import { PaginationPage } from "../browse/_components/Pagination";
+import { assertExists } from "../../lib/utils";
+
+type Room = components["schemas"]["Room"];
 
 interface SelectOption {
   label: string;
@@ -55,6 +68,8 @@ const roomInitialState = {
   recordingType: "cloud",
   recordingTrigger: "automatic-2nd-participant",
   isShared: false,
+  webhookUrl: "",
+  webhookSecret: "",
 };
 
 export default function RoomsList() {
@@ -72,61 +87,77 @@ export default function RoomsList() {
   const recordingTypeCollection = createListCollection({
     items: recordingTypeOptions,
   });
-  const [room, setRoom] = useState(roomInitialState);
+  const [roomInput, setRoomInput] = useState<null | typeof roomInitialState>(
+    null,
+  );
   const [isEditing, setIsEditing] = useState(false);
-  const [editRoomId, setEditRoomId] = useState("");
-  const api = useApi();
-  // TODO seems to be no setPage calls
-  const [page, setPage] = useState<number>(1);
-  const { loading, response, refetch } = useRoomList(PaginationPage(page));
-  const [streams, setStreams] = useState<Stream[]>([]);
-  const [topics, setTopics] = useState<Topic[]>([]);
+  const [editRoomId, setEditRoomId] = useState<string | null>(null);
+  const {
+    loading,
+    response,
+    refetch,
+    error: roomListError,
+  } = useRoomList(PaginationPage(1));
   const [nameError, setNameError] = useState("");
   const [linkCopied, setLinkCopied] = useState("");
-  interface Stream {
-    stream_id: number;
-    name: string;
-  }
+  const [selectedStreamId, setSelectedStreamId] = useState<number | null>(null);
+  const [testingWebhook, setTestingWebhook] = useState(false);
+  const [webhookTestResult, setWebhookTestResult] = useState<string | null>(
+    null,
+  );
+  const [showWebhookSecret, setShowWebhookSecret] = useState(false);
 
-  interface Topic {
-    name: string;
-  }
+  const createRoomMutation = useRoomCreate();
+  const updateRoomMutation = useRoomUpdate();
+  const deleteRoomMutation = useRoomDelete();
+  const { data: streams = [] } = useZulipStreams();
+  const { data: topics = [] } = useZulipTopics(selectedStreamId);
 
+  const {
+    data: detailedEditedRoom,
+    isLoading: isDetailedEditedRoomLoading,
+    error: detailedEditedRoomError,
+  } = useRoomGet(editRoomId);
+
+  const error = roomListError || detailedEditedRoomError;
+
+  // room being edited, as fetched from the server
+  const editedRoom: typeof roomInitialState | null = useMemo(
+    () =>
+      detailedEditedRoom
+        ? {
+            name: detailedEditedRoom.name,
+            zulipAutoPost: detailedEditedRoom.zulip_auto_post,
+            zulipStream: detailedEditedRoom.zulip_stream,
+            zulipTopic: detailedEditedRoom.zulip_topic,
+            isLocked: detailedEditedRoom.is_locked,
+            roomMode: detailedEditedRoom.room_mode,
+            recordingType: detailedEditedRoom.recording_type,
+            recordingTrigger: detailedEditedRoom.recording_trigger,
+            isShared: detailedEditedRoom.is_shared,
+            webhookUrl: detailedEditedRoom.webhook_url || "",
+            webhookSecret: detailedEditedRoom.webhook_secret || "",
+          }
+        : null,
+    [detailedEditedRoom],
+  );
+
+  // a room input value or a last api room state
+  const room = roomInput || editedRoom || roomInitialState;
+
+  const roomTestWebhookMutation = useRoomTestWebhook();
+
+  // Update selected stream ID when zulip stream changes
   useEffect(() => {
-    const fetchZulipStreams = async () => {
-      if (!api) return;
-
-      try {
-        const response = await api.v1ZulipGetStreams();
-        setStreams(response);
-      } catch (error) {
-        console.error("Error fetching Zulip streams:", error);
+    if (room.zulipStream && streams.length > 0) {
+      const selectedStream = streams.find((s) => s.name === room.zulipStream);
+      if (selectedStream !== undefined) {
+        setSelectedStreamId(selectedStream.stream_id);
       }
-    };
-
-    if (room.zulipAutoPost) {
-      fetchZulipStreams();
+    } else {
+      setSelectedStreamId(null);
     }
-  }, [room.zulipAutoPost, !api]);
-
-  useEffect(() => {
-    const fetchZulipTopics = async () => {
-      if (!api || !room.zulipStream) return;
-      try {
-        const selectedStream = streams.find((s) => s.name === room.zulipStream);
-        if (selectedStream) {
-          const response = await api.v1ZulipGetTopics({
-            streamId: selectedStream.stream_id,
-          });
-          setTopics(response);
-        }
-      } catch (error) {
-        console.error("Error fetching Zulip topics:", error);
-      }
-    };
-
-    fetchZulipTopics();
-  }, [room.zulipStream, streams, api]);
+  }, [room.zulipStream, streams]);
 
   const streamOptions: SelectOption[] = streams.map((stream) => {
     return { label: stream.name, value: stream.name };
@@ -155,6 +186,76 @@ export default function RoomsList() {
     }, 2000);
   };
 
+  const handleCloseDialog = () => {
+    setShowWebhookSecret(false);
+    setWebhookTestResult(null);
+    setEditRoomId(null);
+    onClose();
+  };
+
+  const handleTestWebhook = async () => {
+    if (!room.webhookUrl) {
+      setWebhookTestResult("Please enter a webhook URL first");
+      return;
+    }
+    if (!editRoomId) {
+      console.error("No room ID to test webhook");
+      return;
+    }
+
+    setTestingWebhook(true);
+    setWebhookTestResult(null);
+
+    try {
+      const response = await roomTestWebhookMutation.mutateAsync({
+        params: {
+          path: {
+            room_id: editRoomId,
+          },
+        },
+      });
+
+      if (response.success) {
+        setWebhookTestResult(
+          `✅ Webhook test successful! Status: ${response.status_code}`,
+        );
+      } else {
+        let errorMsg = `❌ Webhook test failed`;
+        errorMsg += ` (Status: ${response.status_code})`;
+        if (response.error) {
+          errorMsg += `: ${response.error}`;
+        } else if (response.response_preview) {
+          // Try to parse and extract meaningful error from response
+          // Specific to N8N at the moment, as there is no specification for that
+          // We could just display as is, but decided here to dig a little bit more.
+          try {
+            const preview = JSON.parse(response.response_preview);
+            if (preview.message) {
+              errorMsg += `: ${preview.message}`;
+            }
+          } catch {
+            // If not JSON, just show the preview text (truncated)
+            const previewText = response.response_preview.substring(0, 150);
+            errorMsg += `: ${previewText}`;
+          }
+        } else if (response?.message) {
+          errorMsg += `: ${response.message}`;
+        }
+        setWebhookTestResult(errorMsg);
+      }
+    } catch (error) {
+      console.error("Error testing webhook:", error);
+      setWebhookTestResult("❌ Failed to test webhook. Please check your URL.");
+    } finally {
+      setTestingWebhook(false);
+    }
+
+    // Clear result after 5 seconds
+    setTimeout(() => {
+      setWebhookTestResult(null);
+    }, 5000);
+  };
+
   const handleSaveRoom = async () => {
     try {
       if (RESERVED_PATHS.includes(room.name)) {
@@ -172,30 +273,34 @@ export default function RoomsList() {
         recording_type: room.recordingType,
         recording_trigger: room.recordingTrigger,
         is_shared: room.isShared,
+        webhook_url: room.webhookUrl,
+        webhook_secret: room.webhookSecret,
       };
 
       if (isEditing) {
-        await api?.v1RoomsUpdate({
-          roomId: editRoomId,
-          requestBody: roomData,
+        await updateRoomMutation.mutateAsync({
+          params: {
+            path: { room_id: assertExists(editRoomId) },
+          },
+          body: roomData,
         });
       } else {
-        await api?.v1RoomsCreate({
-          requestBody: roomData,
+        await createRoomMutation.mutateAsync({
+          body: roomData,
         });
       }
 
-      setRoom(roomInitialState);
+      setRoomInput(null);
       setIsEditing(false);
       setEditRoomId("");
       setNameError("");
       refetch();
       onClose();
-    } catch (err) {
+      handleCloseDialog();
+    } catch (err: any) {
       if (
-        err instanceof ApiError &&
-        err.status === 400 &&
-        (err.body as any).detail == "Room name is not unique"
+        err?.status === 400 &&
+        err?.body?.detail == "Room name is not unique"
       ) {
         setNameError(
           "This room name is already taken. Please choose a different name.",
@@ -206,18 +311,11 @@ export default function RoomsList() {
     }
   };
 
-  const handleEditRoom = (roomId, roomData) => {
-    setRoom({
-      name: roomData.name,
-      zulipAutoPost: roomData.zulip_auto_post,
-      zulipStream: roomData.zulip_stream,
-      zulipTopic: roomData.zulip_topic,
-      isLocked: roomData.is_locked,
-      roomMode: roomData.room_mode,
-      recordingType: roomData.recording_type,
-      recordingTrigger: roomData.recording_trigger,
-      isShared: roomData.is_shared,
-    });
+  const handleEditRoom = async (roomId: string, roomData) => {
+    // Reset states
+    setShowWebhookSecret(false);
+    setWebhookTestResult(null);
+
     setEditRoomId(roomId);
     setIsEditing(true);
     setNameError("");
@@ -226,8 +324,10 @@ export default function RoomsList() {
 
   const handleDeleteRoom = async (roomId: string) => {
     try {
-      await api?.v1RoomsDelete({
-        roomId,
+      await deleteRoomMutation.mutateAsync({
+        params: {
+          path: { room_id: roomId },
+        },
       });
       refetch();
     } catch (err) {
@@ -244,7 +344,7 @@ export default function RoomsList() {
         .toLowerCase();
       setNameError("");
     }
-    setRoom({
+    setRoomInput({
       ...room,
       [name]: type === "checkbox" ? checked : value,
     });
@@ -267,6 +367,9 @@ export default function RoomsList() {
       </Flex>
     );
 
+  if (roomListError)
+    return <div>{`${roomListError.name}: ${roomListError.message}`}</div>;
+
   return (
     <Flex
       flexDir="column"
@@ -285,8 +388,10 @@ export default function RoomsList() {
           colorPalette="primary"
           onClick={() => {
             setIsEditing(false);
-            setRoom(roomInitialState);
+            setRoomInput(null);
             setNameError("");
+            setShowWebhookSecret(false);
+            setWebhookTestResult(null);
             onOpen();
           }}
         >
@@ -296,7 +401,7 @@ export default function RoomsList() {
 
       <Dialog.Root
         open={open}
-        onOpenChange={(e) => (e.open ? onOpen() : onClose())}
+        onOpenChange={(e) => (e.open ? onOpen() : handleCloseDialog())}
         size="lg"
       >
         <Dialog.Backdrop />
@@ -352,7 +457,7 @@ export default function RoomsList() {
                 <Select.Root
                   value={[room.roomMode]}
                   onValueChange={(e) =>
-                    setRoom({ ...room, roomMode: e.value[0] })
+                    setRoomInput({ ...room, roomMode: e.value[0] })
                   }
                   collection={roomModeCollection}
                 >
@@ -382,7 +487,7 @@ export default function RoomsList() {
                 <Select.Root
                   value={[room.recordingType]}
                   onValueChange={(e) =>
-                    setRoom({
+                    setRoomInput({
                       ...room,
                       recordingType: e.value[0],
                       recordingTrigger:
@@ -417,7 +522,7 @@ export default function RoomsList() {
                 <Select.Root
                   value={[room.recordingTrigger]}
                   onValueChange={(e) =>
-                    setRoom({ ...room, recordingTrigger: e.value[0] })
+                    setRoomInput({ ...room, recordingTrigger: e.value[0] })
                   }
                   collection={recordingTriggerCollection}
                   disabled={room.recordingType !== "cloud"}
@@ -472,7 +577,7 @@ export default function RoomsList() {
                 <Select.Root
                   value={room.zulipStream ? [room.zulipStream] : []}
                   onValueChange={(e) =>
-                    setRoom({
+                    setRoomInput({
                       ...room,
                       zulipStream: e.value[0],
                       zulipTopic: "",
@@ -507,7 +612,7 @@ export default function RoomsList() {
                 <Select.Root
                   value={room.zulipTopic ? [room.zulipTopic] : []}
                   onValueChange={(e) =>
-                    setRoom({ ...room, zulipTopic: e.value[0] })
+                    setRoomInput({ ...room, zulipTopic: e.value[0] })
                   }
                   collection={topicCollection}
                   disabled={!room.zulipAutoPost}
@@ -533,6 +638,109 @@ export default function RoomsList() {
                   </Select.Positioner>
                 </Select.Root>
               </Field.Root>
+
+              {/* Webhook Configuration Section */}
+              <Field.Root mt={8}>
+                <Field.Label>Webhook URL</Field.Label>
+                <Input
+                  name="webhookUrl"
+                  type="url"
+                  placeholder="https://example.com/webhook"
+                  value={room.webhookUrl}
+                  onChange={handleRoomChange}
+                />
+                <Field.HelperText>
+                  Optional: URL to receive notifications when transcripts are
+                  ready
+                </Field.HelperText>
+              </Field.Root>
+
+              {room.webhookUrl && (
+                <>
+                  <Field.Root mt={4}>
+                    <Field.Label>Webhook Secret</Field.Label>
+                    <Flex gap={2}>
+                      <Input
+                        name="webhookSecret"
+                        type={showWebhookSecret ? "text" : "password"}
+                        value={room.webhookSecret}
+                        onChange={handleRoomChange}
+                        placeholder={
+                          isEditing && room.webhookSecret
+                            ? "••••••••"
+                            : "Leave empty to auto-generate"
+                        }
+                        flex="1"
+                      />
+                      {isEditing && room.webhookSecret && (
+                        <IconButton
+                          size="sm"
+                          variant="ghost"
+                          aria-label={
+                            showWebhookSecret ? "Hide secret" : "Show secret"
+                          }
+                          onClick={() =>
+                            setShowWebhookSecret(!showWebhookSecret)
+                          }
+                        >
+                          {showWebhookSecret ? <LuEyeOff /> : <LuEye />}
+                        </IconButton>
+                      )}
+                    </Flex>
+                    <Field.HelperText>
+                      Used for HMAC signature verification (auto-generated if
+                      left empty)
+                    </Field.HelperText>
+                  </Field.Root>
+
+                  {isEditing && (
+                    <>
+                      <Flex
+                        mt={2}
+                        gap={2}
+                        alignItems="flex-start"
+                        direction="column"
+                      >
+                        <Button
+                          size="sm"
+                          variant="outline"
+                          onClick={handleTestWebhook}
+                          disabled={testingWebhook || !room.webhookUrl}
+                        >
+                          {testingWebhook ? (
+                            <>
+                              <Spinner size="xs" mr={2} />
+                              Testing...
+                            </>
+                          ) : (
+                            "Test Webhook"
+                          )}
+                        </Button>
+                        {webhookTestResult && (
+                          <div
+                            style={{
+                              fontSize: "14px",
+                              wordBreak: "break-word",
+                              maxWidth: "100%",
+                              padding: "8px",
+                              borderRadius: "4px",
+                              backgroundColor: webhookTestResult.startsWith(
+                                "✅",
+                              )
+                                ? "#f0fdf4"
+                                : "#fef2f2",
+                              border: `1px solid ${webhookTestResult.startsWith("✅") ? "#86efac" : "#fca5a5"}`,
+                            }}
+                          >
+                            {webhookTestResult}
+                          </div>
+                        )}
+                      </Flex>
+                    </>
+                  )}
+                </>
+              )}
+
               <Field.Root mt={4}>
                 <Checkbox.Root
                   name="isShared"
@@ -557,7 +765,7 @@ export default function RoomsList() {
               </Field.Root>
             </Dialog.Body>
             <Dialog.Footer>
-              <Button variant="ghost" onClick={onClose}>
+              <Button variant="ghost" onClick={handleCloseDialog}>
                 Cancel
               </Button>
               <Button

www/app/(app)/rooms/useRoomList.tsx

@@ -1,7 +1,7 @@
-import { useEffect, useState } from "react";
-import { useError } from "../../(errors)/errorContext";
-import useApi from "../../lib/useApi";
-import { Page_Room_ } from "../../api";
+import { useRoomsList } from "../../lib/apiHooks";
+import type { components } from "../../reflector-api";
+
+type Page_Room_ = components["schemas"]["Page_RoomDetails_"];
 import { PaginationPage } from "../browse/_components/Pagination";
 
 type RoomList = {
@@ -11,38 +11,17 @@ type RoomList = {
   refetch: () => void;
 };
 
-//always protected
+// Wrapper to maintain backward compatibility
 const useRoomList = (page: PaginationPage): RoomList => {
-  const [response, setResponse] = useState<Page_Room_ | null>(null);
-  const [loading, setLoading] = useState<boolean>(true);
-  const [error, setErrorState] = useState<Error | null>(null);
-  const { setError } = useError();
-  const api = useApi();
-  const [refetchCount, setRefetchCount] = useState(0);
-
-  const refetch = () => {
-    setLoading(true);
-    setRefetchCount(refetchCount + 1);
+  const { data, isLoading, error, refetch } = useRoomsList(page);
+  return {
+    response: data || null,
+    loading: isLoading,
+    error: error
+      ? new Error(error.detail ? JSON.stringify(error.detail) : undefined)
+      : null,
+    refetch,
   };
-
-  useEffect(() => {
-    if (!api) return;
-    setLoading(true);
-    api
-      .v1RoomsList({ page })
-      .then((response) => {
-        setResponse(response);
-        setLoading(false);
-      })
-      .catch((err) => {
-        setResponse(null);
-        setLoading(false);
-        setError(err);
-        setErrorState(err);
-      });
-  }, [!api, page, refetchCount]);
-
-  return { response, loading, error, refetch };
 };
 
 export default useRoomList;

www/app/(app)/transcripts/[transcriptId]/correct/page.tsx

@@ -6,9 +6,10 @@ import TopicPlayer from "./topicPlayer";
 import useParticipants from "../../useParticipants";
 import useTopicWithWords from "../../useTopicWithWords";
 import ParticipantList from "./participantList";
-import { GetTranscriptTopic } from "../../../../api";
+import type { components } from "../../../../reflector-api";
+type GetTranscriptTopic = components["schemas"]["GetTranscriptTopic"];
 import { SelectedText, selectedTextIsTimeSlice } from "./types";
-import useApi from "../../../../lib/useApi";
+import { useTranscriptUpdate } from "../../../../lib/apiHooks";
 import useTranscript from "../../useTranscript";
 import { useError } from "../../../../(errors)/errorContext";
 import { useRouter } from "next/navigation";
@@ -23,7 +24,7 @@ export type TranscriptCorrect = {
 export default function TranscriptCorrect({
   params: { transcriptId },
 }: TranscriptCorrect) {
-  const api = useApi();
+  const updateTranscriptMutation = useTranscriptUpdate();
   const transcript = useTranscript(transcriptId);
   const stateCurrentTopic = useState<GetTranscriptTopic>();
   const [currentTopic, _sct] = stateCurrentTopic;
@@ -34,16 +35,21 @@ export default function TranscriptCorrect({
   const { setError } = useError();
   const router = useRouter();
 
-  const markAsDone = () => {
+  const markAsDone = async () => {
     if (transcript.response && !transcript.response.reviewed) {
-      api
-        ?.v1TranscriptUpdate({ transcriptId, requestBody: { reviewed: true } })
-        .then(() => {
-          router.push(`/transcripts/${transcriptId}`);
-        })
-        .catch((e) => {
-          setError(e, "Error marking as done");
+      try {
+        await updateTranscriptMutation.mutateAsync({
+          params: {
+            path: {
+              transcript_id: transcriptId,
+            },
+          },
+          body: { reviewed: true },
         });
+        router.push(`/transcripts/${transcriptId}`);
+      } catch (e) {
+        setError(e as Error, "Error marking as done");
+      }
     }
   };
 

www/app/(app)/transcripts/[transcriptId]/correct/participantList.tsx

@@ -1,8 +1,15 @@
 import { faArrowTurnDown } from "@fortawesome/free-solid-svg-icons";
 import { FontAwesomeIcon } from "@fortawesome/react-fontawesome";
 import { ChangeEvent, useEffect, useRef, useState } from "react";
-import { Participant } from "../../../../api";
-import useApi from "../../../../lib/useApi";
+import type { components } from "../../../../reflector-api";
+type Participant = components["schemas"]["Participant"];
+import {
+  useTranscriptSpeakerAssign,
+  useTranscriptSpeakerMerge,
+  useTranscriptParticipantUpdate,
+  useTranscriptParticipantCreate,
+  useTranscriptParticipantDelete,
+} from "../../../../lib/apiHooks";
 import { UseParticipants } from "../../useParticipants";
 import { selectedTextIsSpeaker, selectedTextIsTimeSlice } from "./types";
 import { useError } from "../../../../(errors)/errorContext";
@@ -30,9 +37,19 @@ const ParticipantList = ({
   topicWithWords,
   stateSelectedText,
 }: ParticipantList) => {
-  const api = useApi();
   const { setError } = useError();
-  const [loading, setLoading] = useState(false);
+  const speakerAssignMutation = useTranscriptSpeakerAssign();
+  const speakerMergeMutation = useTranscriptSpeakerMerge();
+  const participantUpdateMutation = useTranscriptParticipantUpdate();
+  const participantCreateMutation = useTranscriptParticipantCreate();
+  const participantDeleteMutation = useTranscriptParticipantDelete();
+
+  const loading =
+    speakerAssignMutation.isPending ||
+    speakerMergeMutation.isPending ||
+    participantUpdateMutation.isPending ||
+    participantCreateMutation.isPending ||
+    participantDeleteMutation.isPending;
   const [participantInput, setParticipantInput] = useState("");
   const inputRef = useRef<HTMLInputElement>(null);
   const [selectedText, setSelectedText] = stateSelectedText;
@@ -103,7 +120,6 @@ const ParticipantList = ({
   const onSuccess = () => {
     topicWithWords.refetch();
     participants.refetch();
-    setLoading(false);
     setAction(null);
     setSelectedText(undefined);
     setSelectedParticipant(undefined);
@@ -120,11 +136,14 @@ const ParticipantList = ({
       if (loading || participants.loading || topicWithWords.loading) return;
       if (!selectedTextIsTimeSlice(selectedText)) return;
 
-      setLoading(true);
       try {
-        await api?.v1TranscriptAssignSpeaker({
-          transcriptId,
-          requestBody: {
+        await speakerAssignMutation.mutateAsync({
+          params: {
+            path: {
+              transcript_id: transcriptId,
+            },
+          },
+          body: {
             participant: participant.id,
             timestamp_from: selectedText.start,
             timestamp_to: selectedText.end,
@@ -132,8 +151,7 @@ const ParticipantList = ({
         });
         onSuccess();
       } catch (error) {
-        setError(error, "There was an error assigning");
-        setLoading(false);
+        setError(error as Error, "There was an error assigning");
         throw error;
       }
     };
@@ -141,32 +159,38 @@ const ParticipantList = ({
   const mergeSpeaker =
     (speakerFrom, participantTo: Participant) => async () => {
       if (loading || participants.loading || topicWithWords.loading) return;
-      setLoading(true);
+
       if (participantTo.speaker) {
         try {
-          await api?.v1TranscriptMergeSpeaker({
-            transcriptId,
-            requestBody: {
+          await speakerMergeMutation.mutateAsync({
+            params: {
+              path: {
+                transcript_id: transcriptId,
+              },
+            },
+            body: {
               speaker_from: speakerFrom,
               speaker_to: participantTo.speaker,
             },
           });
           onSuccess();
         } catch (error) {
-          setError(error, "There was an error merging");
-          setLoading(false);
+          setError(error as Error, "There was an error merging");
         }
       } else {
         try {
-          await api?.v1TranscriptUpdateParticipant({
-            transcriptId,
-            participantId: participantTo.id,
-            requestBody: { speaker: speakerFrom },
+          await participantUpdateMutation.mutateAsync({
+            params: {
+              path: {
+                transcript_id: transcriptId,
+                participant_id: participantTo.id,
+              },
+            },
+            body: { speaker: speakerFrom },
           });
           onSuccess();
         } catch (error) {
-          setError(error, "There was an error merging (update)");
-          setLoading(false);
+          setError(error as Error, "There was an error merging (update)");
         }
       }
     };
@@ -186,105 +210,106 @@ const ParticipantList = ({
         (p) => p.speaker == selectedText,
       );
       if (participant && participant.name !== participantInput) {
-        setLoading(true);
-        api
-          ?.v1TranscriptUpdateParticipant({
-            transcriptId,
-            participantId: participant.id,
-            requestBody: {
+        try {
+          await participantUpdateMutation.mutateAsync({
+            params: {
+              path: {
+                transcript_id: transcriptId,
+                participant_id: participant.id,
+              },
+            },
+            body: {
               name: participantInput,
             },
-          })
-          .then(() => {
-            participants.refetch();
-            setLoading(false);
-            setAction(null);
-          })
-          .catch((e) => {
-            setError(e, "There was an error renaming");
-            setLoading(false);
           });
+          participants.refetch();
+          setAction(null);
+        } catch (e) {
+          setError(e as Error, "There was an error renaming");
+        }
       }
     } else if (
       action == "Create to rename" &&
       selectedTextIsSpeaker(selectedText)
     ) {
-      setLoading(true);
-      api
-        ?.v1TranscriptAddParticipant({
-          transcriptId,
-          requestBody: {
+      try {
+        await participantCreateMutation.mutateAsync({
+          params: {
+            path: {
+              transcript_id: transcriptId,
+            },
+          },
+          body: {
             name: participantInput,
             speaker: selectedText,
           },
-        })
-        .then(() => {
-          participants.refetch();
-          setParticipantInput("");
-          setOneMatch(undefined);
-          setLoading(false);
-        })
-        .catch((e) => {
-          setError(e, "There was an error creating");
-          setLoading(false);
         });
+        participants.refetch();
+        setParticipantInput("");
+        setOneMatch(undefined);
+      } catch (e) {
+        setError(e as Error, "There was an error creating");
+      }
     } else if (
       action == "Create and assign" &&
       selectedTextIsTimeSlice(selectedText)
     ) {
-      setLoading(true);
       try {
-        const participant = await api?.v1TranscriptAddParticipant({
-          transcriptId,
-          requestBody: {
+        const participant = await participantCreateMutation.mutateAsync({
+          params: {
+            path: {
+              transcript_id: transcriptId,
+            },
+          },
+          body: {
             name: participantInput,
           },
         });
-        setLoading(false);
         assignTo(participant)().catch(() => {
           // error and loading are handled by assignTo catch
           participants.refetch();
         });
       } catch (error) {
-        setError(e, "There was an error creating");
-        setLoading(false);
+        setError(error as Error, "There was an error creating");
       }
     } else if (action == "Create") {
-      setLoading(true);
-      api
-        ?.v1TranscriptAddParticipant({
-          transcriptId,
-          requestBody: {
+      try {
+        await participantCreateMutation.mutateAsync({
+          params: {
+            path: {
+              transcript_id: transcriptId,
+            },
+          },
+          body: {
             name: participantInput,
           },
-        })
-        .then(() => {
-          participants.refetch();
-          setParticipantInput("");
-          setLoading(false);
-          inputRef.current?.focus();
-        })
-        .catch((e) => {
-          setError(e, "There was an error creating");
-          setLoading(false);
         });
+        participants.refetch();
+        setParticipantInput("");
+        inputRef.current?.focus();
+      } catch (e) {
+        setError(e as Error, "There was an error creating");
+      }
     }
   };
 
-  const deleteParticipant = (participantId) => (e) => {
+  const deleteParticipant = (participantId) => async (e) => {
     e.stopPropagation();
     if (loading || participants.loading || topicWithWords.loading) return;
-    setLoading(true);
-    api
-      ?.v1TranscriptDeleteParticipant({ transcriptId, participantId })
-      .then(() => {
-        participants.refetch();
-        setLoading(false);
-      })
-      .catch((e) => {
-        setError(e, "There was an error deleting");
-        setLoading(false);
+
+    try {
+      await participantDeleteMutation.mutateAsync({
+        params: {
+          path: {
+            transcript_id: transcriptId,
+            participant_id: participantId,
+          },
+        },
       });
+      participants.refetch();
+    } catch (e) {
+      setError(e as Error, "There was an error deleting");
+    }
   };
 
   const selectParticipant = (participant) => (e) => {

www/app/(app)/transcripts/[transcriptId]/correct/topicHeader.tsx

@@ -1,6 +1,7 @@
 import useTopics from "../../useTopics";
 import { Dispatch, SetStateAction, useEffect } from "react";
-import { GetTranscriptTopic } from "../../../../api";
+import type { components } from "../../../../reflector-api";
+type GetTranscriptTopic = components["schemas"]["GetTranscriptTopic"];
 import {
   BoxProps,
   Box,

www/app/(app)/transcripts/[transcriptId]/finalSummary.tsx

@@ -2,12 +2,10 @@ import { useEffect, useRef, useState } from "react";
 import React from "react";
 import Markdown from "react-markdown";
 import "../../../styles/markdown.css";
-import {
-  GetTranscript,
-  GetTranscriptTopic,
-  UpdateTranscript,
-} from "../../../api";
-import useApi from "../../../lib/useApi";
+import type { components } from "../../../reflector-api";
+type GetTranscript = components["schemas"]["GetTranscript"];
+type GetTranscriptTopic = components["schemas"]["GetTranscriptTopic"];
+import { useTranscriptUpdate } from "../../../lib/apiHooks";
 import {
   Flex,
   Heading,
@@ -33,9 +31,8 @@ export default function FinalSummary(props: FinalSummaryProps) {
   const [preEditSummary, setPreEditSummary] = useState("");
   const [editedSummary, setEditedSummary] = useState("");
 
-  const api = useApi();
-
   const { setError } = useError();
+  const updateTranscriptMutation = useTranscriptUpdate();
 
   useEffect(() => {
     setEditedSummary(props.transcriptResponse?.long_summary || "");
@@ -47,12 +44,15 @@ export default function FinalSummary(props: FinalSummaryProps) {
 
   const updateSummary = async (newSummary: string, transcriptId: string) => {
     try {
-      const requestBody: UpdateTranscript = {
-        long_summary: newSummary,
-      };
-      const updatedTranscript = await api?.v1TranscriptUpdate({
-        transcriptId,
-        requestBody,
+      const updatedTranscript = await updateTranscriptMutation.mutateAsync({
+        params: {
+          path: {
+            transcript_id: transcriptId,
+          },
+        },
+        body: {
+          long_summary: newSummary,
+        },
       });
       if (props.onUpdate) {
         props.onUpdate(newSummary);
@@ -60,7 +60,7 @@ export default function FinalSummary(props: FinalSummaryProps) {
       console.log("Updated long summary:", updatedTranscript);
     } catch (err) {
       console.error("Failed to update long summary:", err);
-      setError(err, "Failed to update long summary.");
+      setError(err as Error, "Failed to update long summary.");
     }
   };
 
@@ -114,7 +114,12 @@ export default function FinalSummary(props: FinalSummaryProps) {
             <Button onClick={onDiscardClick} variant="ghost">
               Cancel
             </Button>
-            <Button onClick={onSaveClick}>Save</Button>
+            <Button
+              onClick={onSaveClick}
+              disabled={updateTranscriptMutation.isPending}
+            >
+              Save
+            </Button>
           </Flex>
         )}
         {!isEditMode && (

www/app/(app)/transcripts/[transcriptId]/page.tsx

@@ -86,7 +86,7 @@ export default function TranscriptDetails(details: TranscriptDetails) {
                 useActiveTopic={useActiveTopic}
                 waveform={waveform.waveform}
                 media={mp3.media}
-                mediaDuration={transcript.response.duration}
+                mediaDuration={transcript.response?.duration || null}
               />
             ) : !mp3.loading && (waveform.error || mp3.error) ? (
               <Box p={4} bg="red.100" borderRadius="md">
@@ -116,7 +116,7 @@ export default function TranscriptDetails(details: TranscriptDetails) {
             <Flex direction="column" gap={0}>
               <Flex alignItems="center" gap={2}>
                 <TranscriptTitle
-                  title={transcript.response.title || "Unnamed Transcript"}
+                  title={transcript.response?.title || "Unnamed Transcript"}
                   transcriptId={transcriptId}
                   onUpdate={(newTitle) => {
                     transcript.reload();

www/app/(app)/transcripts/[transcriptId]/upload/page.tsx

@@ -24,10 +24,16 @@ const TranscriptUpload = (details: TranscriptUpload) => {
 
   const router = useRouter();
 
-  const [status, setStatus] = useState(
+  const [status_, setStatus] = useState(
     webSockets.status.value || transcript.response?.status || "idle",
   );
 
+  // status is obviously done if we have transcript
+  const status =
+    !transcript.loading && transcript.response?.status === "ended"
+      ? transcript.response?.status
+      : status_;
+
   useEffect(() => {
     if (!transcriptStarted && webSockets.transcriptTextLive.length !== 0)
       setTranscriptStarted(true);
@@ -35,8 +41,11 @@ const TranscriptUpload = (details: TranscriptUpload) => {
 
   useEffect(() => {
     //TODO HANDLE ERROR STATUS BETTER
+    // TODO deprecate webSockets.status.value / depend on transcript.response?.status from query lib
     const newStatus =
-      webSockets.status.value || transcript.response?.status || "idle";
+      transcript.response?.status === "ended"
+        ? "ended"
+        : webSockets.status.value || transcript.response?.status || "idle";
     setStatus(newStatus);
     if (newStatus && (newStatus == "ended" || newStatus == "error")) {
       console.log(newStatus, "redirecting");

www/app/(app)/transcripts/createTranscript.ts

@@ -1,45 +1,33 @@
-import { useEffect, useState } from "react";
+import type { components } from "../../reflector-api";
+import { useTranscriptCreate } from "../../lib/apiHooks";
 
-import { useError } from "../../(errors)/errorContext";
-import { CreateTranscript, GetTranscript } from "../../api";
-import useApi from "../../lib/useApi";
+type CreateTranscript = components["schemas"]["CreateTranscript"];
+type GetTranscript = components["schemas"]["GetTranscript"];
 
 type UseCreateTranscript = {
   transcript: GetTranscript | null;
   loading: boolean;
   error: Error | null;
-  create: (transcriptCreationDetails: CreateTranscript) => void;
+  create: (transcriptCreationDetails: CreateTranscript) => Promise<void>;
 };
 
 const useCreateTranscript = (): UseCreateTranscript => {
-  const [transcript, setTranscript] = useState<GetTranscript | null>(null);
-  const [loading, setLoading] = useState<boolean>(false);
-  const [error, setErrorState] = useState<Error | null>(null);
-  const { setError } = useError();
-  const api = useApi();
+  const createMutation = useTranscriptCreate();
 
-  const create = (transcriptCreationDetails: CreateTranscript) => {
-    if (loading || !api) return;
+  const create = async (transcriptCreationDetails: CreateTranscript) => {
+    if (createMutation.isPending) return;
 
-    setLoading(true);
-
-    api
-      .v1TranscriptsCreate({ requestBody: transcriptCreationDetails })
-      .then((transcript) => {
-        setTranscript(transcript);
-        setLoading(false);
-      })
-      .catch((err) => {
-        setError(
-          err,
-          "There was an issue creating a transcript, please try again.",
-        );
-        setErrorState(err);
-        setLoading(false);
-      });
+    await createMutation.mutateAsync({
+      body: transcriptCreationDetails,
+    });
   };
 
-  return { transcript, loading, error, create };
+  return {
+    transcript: createMutation.data || null,
+    loading: createMutation.isPending,
+    error: createMutation.error as Error | null,
+    create,
+  };
 };
 
 export default useCreateTranscript;

www/app/(app)/transcripts/fileUploadButton.tsx

@@ -1,6 +1,7 @@
 import React, { useState } from "react";
-import useApi from "../../lib/useApi";
+import { useTranscriptUploadAudio } from "../../lib/apiHooks";
 import { Button, Spinner } from "@chakra-ui/react";
+import { useError } from "../../(errors)/errorContext";
 
 type FileUploadButton = {
   transcriptId: string;
@@ -8,13 +9,16 @@ type FileUploadButton = {
 
 export default function FileUploadButton(props: FileUploadButton) {
   const fileInputRef = React.useRef<HTMLInputElement>(null);
-  const api = useApi();
+  const uploadMutation = useTranscriptUploadAudio();
+  const { setError } = useError();
   const [progress, setProgress] = useState(0);
   const triggerFileUpload = () => {
     fileInputRef.current?.click();
   };
 
-  const handleFileUpload = (event: React.ChangeEvent<HTMLInputElement>) => {
+  const handleFileUpload = async (
+    event: React.ChangeEvent<HTMLInputElement>,
+  ) => {
     const file = event.target.files?.[0];
 
     if (file) {
@@ -24,37 +28,45 @@ export default function FileUploadButton(props: FileUploadButton) {
       let start = 0;
       let uploadedSize = 0;
 
-      api?.httpRequest.config.interceptors.request.use((request) => {
-        request.onUploadProgress = (progressEvent) => {
-          const currentProgress = Math.floor(
-            ((uploadedSize + progressEvent.loaded) / file.size) * 100,
-          );
-          setProgress(currentProgress);
-        };
-        return request;
-      });
-
       const uploadNextChunk = async () => {
-        if (chunkNumber == totalChunks) return;
+        if (chunkNumber == totalChunks) {
+          setProgress(0);
+          return;
+        }
 
         const chunkSize = Math.min(maxChunkSize, file.size - start);
         const end = start + chunkSize;
         const chunk = file.slice(start, end);
 
-        await api?.v1TranscriptRecordUpload({
-          transcriptId: props.transcriptId,
-          formData: {
-            chunk,
-          },
-          chunkNumber,
-          totalChunks,
-        });
+        try {
+          const formData = new FormData();
+          formData.append("chunk", chunk);
 
-        uploadedSize += chunkSize;
-        chunkNumber++;
-        start = end;
+          await uploadMutation.mutateAsync({
+            params: {
+              path: {
+                transcript_id: props.transcriptId,
+              },
+              query: {
+                chunk_number: chunkNumber,
+                total_chunks: totalChunks,
+              },
+            },
+            body: formData as any,
+          });
 
-        uploadNextChunk();
+          uploadedSize += chunkSize;
+          const currentProgress = Math.floor((uploadedSize / file.size) * 100);
+          setProgress(currentProgress);
+
+          chunkNumber++;
+          start = end;
+
+          await uploadNextChunk();
+        } catch (error) {
+          setError(error as Error, "Failed to upload file");
+          setProgress(0);
+        }
       };
 
       uploadNextChunk();

www/app/(app)/transcripts/new/page.tsx

@@ -9,33 +9,27 @@ import { useRouter } from "next/navigation";
 import useCreateTranscript from "../createTranscript";
 import SelectSearch from "react-select-search";
 import { supportedLanguages } from "../../../supportedLanguages";
-import useSessionStatus from "../../../lib/useSessionStatus";
 import { featureEnabled } from "../../../domainContext";
-import { signIn } from "next-auth/react";
 import {
   Flex,
   Box,
   Spinner,
   Heading,
   Button,
-  Card,
   Center,
-  Link,
-  CardBody,
-  Stack,
   Text,
-  Icon,
-  Grid,
-  IconButton,
   Spacer,
-  Menu,
-  Tooltip,
-  Input,
 } from "@chakra-ui/react";
+import { useAuth } from "../../../lib/AuthProvider";
+import type { components } from "../../../reflector-api";
+
 const TranscriptCreate = () => {
   const isClient = typeof window !== "undefined";
   const router = useRouter();
-  const { isLoading, isAuthenticated } = useSessionStatus();
+  const auth = useAuth();
+  const isAuthenticated = auth.status === "authenticated";
+  const isAuthRefreshing = auth.status === "refreshing";
+  const isLoading = auth.status === "loading";
   const requireLogin = featureEnabled("requireLogin");
 
   const [name, setName] = useState<string>("");
@@ -54,20 +48,32 @@ const TranscriptCreate = () => {
   const [loadingUpload, setLoadingUpload] = useState(false);
 
   const getTargetLanguage = () => {
-    if (targetLanguage === "NOTRANSLATION") return;
+    if (targetLanguage === "NOTRANSLATION") return undefined;
     return targetLanguage;
   };
 
   const send = () => {
     if (loadingRecord || createTranscript.loading || permissionDenied) return;
     setLoadingRecord(true);
-    createTranscript.create({ name, target_language: getTargetLanguage() });
+    const targetLang = getTargetLanguage();
+    createTranscript.create({
+      name,
+      source_language: "en",
+      target_language: targetLang || "en",
+      source_kind: "live",
+    });
   };
 
   const uploadFile = () => {
     if (loadingUpload || createTranscript.loading || permissionDenied) return;
     setLoadingUpload(true);
-    createTranscript.create({ name, target_language: getTargetLanguage() });
+    const targetLang = getTargetLanguage();
+    createTranscript.create({
+      name,
+      source_language: "en",
+      target_language: targetLang || "en",
+      source_kind: "file",
+    });
   };
 
   useEffect(() => {
@@ -132,8 +138,8 @@ const TranscriptCreate = () => {
           <Center>
             {isLoading ? (
               <Spinner />
-            ) : requireLogin && !isAuthenticated ? (
-              <Button onClick={() => signIn("authentik")}>Log in</Button>
+            ) : requireLogin && !isAuthenticated && !isAuthRefreshing ? (
+              <Button onClick={() => auth.signIn("authentik")}>Log in</Button>
             ) : (
               <Flex
                 rounded="xl"

www/app/(app)/transcripts/player.tsx

@@ -5,7 +5,9 @@ import RegionsPlugin from "wavesurfer.js/dist/plugins/regions.esm.js";
 
 import { formatTime, formatTimeMs } from "../../lib/time";
 import { Topic } from "./webSocketTypes";
-import { AudioWaveform } from "../../api";
+import type { components } from "../../reflector-api";
+
+type AudioWaveform = components["schemas"]["AudioWaveform"];
 import { waveSurferStyles } from "../../styles/recorder";
 import { Box, Flex, IconButton } from "@chakra-ui/react";
 import { LuPause, LuPlay } from "react-icons/lu";
@@ -18,7 +20,7 @@ type PlayerProps = {
   ];
   waveform: AudioWaveform;
   media: HTMLMediaElement;
-  mediaDuration: number;
+  mediaDuration: number | null;
 };
 
 export default function Player(props: PlayerProps) {
@@ -50,7 +52,9 @@ export default function Player(props: PlayerProps) {
         container: waveformRef.current,
         peaks: [props.waveform.data],
         height: "auto",
-        duration: Math.floor(props.mediaDuration / 1000),
+        duration: props.mediaDuration
+          ? Math.floor(props.mediaDuration / 1000)
+          : undefined,
         media: props.media,
 
         ...waveSurferStyles.playerSettings,

www/app/(app)/transcripts/recorder.tsx

@@ -6,7 +6,6 @@ import RecordPlugin from "../../lib/custom-plugins/record";
 import { formatTime, formatTimeMs } from "../../lib/time";
 import { waveSurferStyles } from "../../styles/recorder";
 import { useError } from "../../(errors)/errorContext";
-import FileUploadButton from "./fileUploadButton";
 import useWebRTC from "./useWebRTC";
 import useAudioDevice from "./useAudioDevice";
 import { Box, Flex, IconButton, Menu, RadioGroup } from "@chakra-ui/react";

www/app/(app)/transcripts/shareAndPrivacy.tsx

@@ -2,7 +2,10 @@ import { useEffect, useState } from "react";
 import { featureEnabled } from "../../domainContext";
 
 import { ShareMode, toShareMode } from "../../lib/shareMode";
-import { GetTranscript, GetTranscriptTopic, UpdateTranscript } from "../../api";
+import type { components } from "../../reflector-api";
+type GetTranscript = components["schemas"]["GetTranscript"];
+type GetTranscriptTopic = components["schemas"]["GetTranscriptTopic"];
+type UpdateTranscript = components["schemas"]["UpdateTranscript"];
 import {
   Box,
   Flex,
@@ -15,12 +18,11 @@ import {
   createListCollection,
 } from "@chakra-ui/react";
 import { LuShare2 } from "react-icons/lu";
-import useApi from "../../lib/useApi";
-import useSessionUser from "../../lib/useSessionUser";
-import { CustomSession } from "../../lib/types";
+import { useTranscriptUpdate } from "../../lib/apiHooks";
 import ShareLink from "./shareLink";
 import ShareCopy from "./shareCopy";
 import ShareZulip from "./shareZulip";
+import { useAuth } from "../../lib/AuthProvider";
 
 type ShareAndPrivacyProps = {
   finalSummaryRef: any;
@@ -50,12 +52,9 @@ export default function ShareAndPrivacy(props: ShareAndPrivacyProps) {
   );
   const [shareLoading, setShareLoading] = useState(false);
   const requireLogin = featureEnabled("requireLogin");
-  const api = useApi();
+  const updateTranscriptMutation = useTranscriptUpdate();
 
   const updateShareMode = async (selectedValue: string) => {
-    if (!api)
-      throw new Error("ShareLink's API should always be ready at this point");
-
     const selectedOption = shareOptionsData.find(
       (option) => option.value === selectedValue,
     );
@@ -67,19 +66,27 @@ export default function ShareAndPrivacy(props: ShareAndPrivacyProps) {
       share_mode: selectedValue as "public" | "semi-private" | "private",
     };
 
-    const updatedTranscript = await api.v1TranscriptUpdate({
-      transcriptId: props.transcriptResponse.id,
-      requestBody,
-    });
-    setShareMode(
-      shareOptionsData.find(
-        (option) => option.value === updatedTranscript.share_mode,
-      ) || shareOptionsData[0],
-    );
-    setShareLoading(false);
+    try {
+      const updatedTranscript = await updateTranscriptMutation.mutateAsync({
+        params: {
+          path: { transcript_id: props.transcriptResponse.id },
+        },
+        body: requestBody,
+      });
+      setShareMode(
+        shareOptionsData.find(
+          (option) => option.value === updatedTranscript.share_mode,
+        ) || shareOptionsData[0],
+      );
+    } catch (err) {
+      console.error("Failed to update share mode:", err);
+    } finally {
+      setShareLoading(false);
+    }
   };
 
-  const userId = useSessionUser().id;
+  const auth = useAuth();
+  const userId = auth.status === "authenticated" ? auth.user?.id : null;
 
   useEffect(() => {
     setIsOwner(!!(requireLogin && userId === props.transcriptResponse.user_id));
@@ -124,7 +131,7 @@ export default function ShareAndPrivacy(props: ShareAndPrivacyProps) {
                       "This transcript is public. Everyone can access it."}
                   </Text>
 
-                  {isOwner && api && (
+                  {isOwner && (
                     <Select.Root
                       key={shareMode.value}
                       value={[shareMode.value || ""]}

www/app/(app)/transcripts/shareCopy.tsx

@@ -1,5 +1,7 @@
 import { useState } from "react";
-import { GetTranscript, GetTranscriptTopic } from "../../api";
+import type { components } from "../../reflector-api";
+type GetTranscript = components["schemas"]["GetTranscript"];
+type GetTranscriptTopic = components["schemas"]["GetTranscriptTopic"];
 import { Button, BoxProps, Box } from "@chakra-ui/react";
 
 type ShareCopyProps = {

www/app/(app)/transcripts/shareZulip.tsx

@@ -1,6 +1,9 @@
 import { useState, useEffect, useMemo } from "react";
 import { featureEnabled } from "../../domainContext";
-import { GetTranscript, GetTranscriptTopic } from "../../api";
+import type { components } from "../../reflector-api";
+
+type GetTranscript = components["schemas"]["GetTranscript"];
+type GetTranscriptTopic = components["schemas"]["GetTranscriptTopic"];
 import {
   BoxProps,
   Button,
@@ -12,12 +15,15 @@ import {
   Checkbox,
   Combobox,
   Spinner,
-  Portal,
   useFilter,
   useListCollection,
 } from "@chakra-ui/react";
 import { TbBrandZulip } from "react-icons/tb";
-import useApi from "../../lib/useApi";
+import {
+  useZulipStreams,
+  useZulipTopics,
+  useTranscriptPostToZulip,
+} from "../../lib/apiHooks";
 
 type ShareZulipProps = {
   transcriptResponse: GetTranscript;
@@ -30,104 +36,75 @@ interface Stream {
   name: string;
 }
 
-interface Topic {
-  name: string;
-}
-
 export default function ShareZulip(props: ShareZulipProps & BoxProps) {
   const [showModal, setShowModal] = useState(false);
   const [stream, setStream] = useState<string | undefined>(undefined);
+  const [selectedStreamId, setSelectedStreamId] = useState<number | null>(null);
   const [topic, setTopic] = useState<string | undefined>(undefined);
   const [includeTopics, setIncludeTopics] = useState(false);
-  const [isLoading, setIsLoading] = useState(true);
-  const [streams, setStreams] = useState<Stream[]>([]);
-  const [topics, setTopics] = useState<Topic[]>([]);
-  const api = useApi();
+
+  const { data: streams = [], isLoading: isLoadingStreams } = useZulipStreams();
+  const { data: topics = [] } = useZulipTopics(selectedStreamId);
+  const postToZulipMutation = useTranscriptPostToZulip();
+
   const { contains } = useFilter({ sensitivity: "base" });
 
-  const {
-    collection: streamItemsCollection,
-    filter: streamItemsFilter,
-    set: streamItemsSet,
-  } = useListCollection({
-    initialItems: [] as { label: string; value: string }[],
-    filter: contains,
-  });
+  const streamItems = useMemo(() => {
+    return streams.map((stream: Stream) => ({
+      label: stream.name,
+      value: stream.name,
+    }));
+  }, [streams]);
 
-  const {
-    collection: topicItemsCollection,
-    filter: topicItemsFilter,
-    set: topicItemsSet,
-  } = useListCollection({
-    initialItems: [] as { label: string; value: string }[],
-    filter: contains,
-  });
+  const topicItems = useMemo(() => {
+    return topics.map(({ name }) => ({
+      label: name,
+      value: name,
+    }));
+  }, [topics]);
 
+  const { collection: streamItemsCollection, filter: streamItemsFilter } =
+    useListCollection({
+      initialItems: streamItems,
+      filter: contains,
+    });
+
+  const { collection: topicItemsCollection, filter: topicItemsFilter } =
+    useListCollection({
+      initialItems: topicItems,
+      filter: contains,
+    });
+
+  // Update selected stream ID when stream changes
   useEffect(() => {
-    const fetchZulipStreams = async () => {
-      if (!api) return;
-
-      try {
-        const response = await api.v1ZulipGetStreams();
-        setStreams(response);
-
-        streamItemsSet(
-          response.map((stream) => ({
-            label: stream.name,
-            value: stream.name,
-          })),
-        );
-
-        setIsLoading(false);
-      } catch (error) {
-        console.error("Error fetching Zulip streams:", error);
-      }
-    };
-
-    fetchZulipStreams();
-  }, [!api]);
-
-  useEffect(() => {
-    const fetchZulipTopics = async () => {
-      if (!api || !stream) return;
-      try {
-        const selectedStream = streams.find((s) => s.name === stream);
-        if (selectedStream) {
-          const response = await api.v1ZulipGetTopics({
-            streamId: selectedStream.stream_id,
-          });
-          setTopics(response);
-          topicItemsSet(
-            response.map((topic) => ({
-              label: topic.name,
-              value: topic.name,
-            })),
-          );
-        } else {
-          topicItemsSet([]);
-        }
-      } catch (error) {
-        console.error("Error fetching Zulip topics:", error);
-      }
-    };
-
-    fetchZulipTopics();
-  }, [stream, streams, api]);
+    if (stream && streams) {
+      const selectedStream = streams.find((s: Stream) => s.name === stream);
+      setSelectedStreamId(selectedStream ? selectedStream.stream_id : null);
+    } else {
+      setSelectedStreamId(null);
+    }
+  }, [stream, streams]);
 
   const handleSendToZulip = async () => {
-    if (!api || !props.transcriptResponse) return;
+    if (!props.transcriptResponse) return;
 
     if (stream && topic) {
       try {
-        await api.v1TranscriptPostToZulip({
-          transcriptId: props.transcriptResponse.id,
-          stream,
-          topic,
-          includeTopics,
+        await postToZulipMutation.mutateAsync({
+          params: {
+            path: {
+              transcript_id: props.transcriptResponse.id,
+            },
+            query: {
+              stream,
+              topic,
+              include_topics: includeTopics,
+            },
+          },
         });
         setShowModal(false);
       } catch (error) {
-        console.log(error);
+        console.error("Error posting to Zulip:", error);
       }
     }
   };
@@ -155,7 +132,7 @@ export default function ShareZulip(props: ShareZulipProps & BoxProps) {
               </Dialog.CloseTrigger>
             </Dialog.Header>
             <Dialog.Body>
-              {isLoading ? (
+              {isLoadingStreams ? (
                 <Flex justify="center" py={8}>
                   <Spinner />
                 </Flex>

www/app/(app)/transcripts/transcriptTitle.tsx

@@ -1,6 +1,8 @@
 import { useState } from "react";
-import { UpdateTranscript } from "../../api";
-import useApi from "../../lib/useApi";
+import type { components } from "../../reflector-api";
+
+type UpdateTranscript = components["schemas"]["UpdateTranscript"];
+import { useTranscriptUpdate } from "../../lib/apiHooks";
 import { Heading, IconButton, Input, Flex, Spacer } from "@chakra-ui/react";
 import { LuPen } from "react-icons/lu";
 
@@ -14,24 +16,27 @@ const TranscriptTitle = (props: TranscriptTitle) => {
   const [displayedTitle, setDisplayedTitle] = useState(props.title);
   const [preEditTitle, setPreEditTitle] = useState(props.title);
   const [isEditing, setIsEditing] = useState(false);
-  const api = useApi();
+  const updateTranscriptMutation = useTranscriptUpdate();
 
   const updateTitle = async (newTitle: string, transcriptId: string) => {
-    if (!api) return;
     try {
       const requestBody: UpdateTranscript = {
         title: newTitle,
       };
-      const updatedTranscript = await api?.v1TranscriptUpdate({
-        transcriptId,
-        requestBody,
+      await updateTranscriptMutation.mutateAsync({
+        params: {
+          path: { transcript_id: transcriptId },
+        },
+        body: requestBody,
       });
       if (props.onUpdate) {
         props.onUpdate(newTitle);
       }
-      console.log("Updated transcript:", updatedTranscript);
+      console.log("Updated transcript title:", newTitle);
     } catch (err) {
       console.error("Failed to update transcript:", err);
+      // Revert title on error
+      setDisplayedTitle(preEditTitle);
     }
   };
 

www/app/(app)/transcripts/useMp3.ts

@@ -1,6 +1,7 @@
 import { useContext, useEffect, useState } from "react";
 import { DomainContext } from "../../domainContext";
-import getApi from "../../lib/useApi";
+import { useTranscriptGet } from "../../lib/apiHooks";
+import { useAuth } from "../../lib/AuthProvider";
 
 export type Mp3Response = {
   media: HTMLMediaElement | null;
@@ -17,14 +18,17 @@ const useMp3 = (transcriptId: string, waiting?: boolean): Mp3Response => {
   const [audioLoadingError, setAudioLoadingError] = useState<null | string>(
     null,
   );
-  const [transcriptMetadataLoading, setTranscriptMetadataLoading] =
-    useState<boolean>(true);
-  const [transcriptMetadataLoadingError, setTranscriptMetadataLoadingError] =
-    useState<string | null>(null);
   const [audioDeleted, setAudioDeleted] = useState<boolean | null>(null);
-  const api = getApi();
   const { api_url } = useContext(DomainContext);
-  const accessTokenInfo = api?.httpRequest?.config?.TOKEN;
+  const auth = useAuth();
+  const accessTokenInfo =
+    auth.status === "authenticated" ? auth.accessToken : null;
+
+  const {
+    data: transcript,
+    isLoading: transcriptMetadataLoading,
+    error: transcriptError,
+  } = useTranscriptGet(later ? null : transcriptId);
 
   const [serviceWorker, setServiceWorker] =
     useState<ServiceWorkerRegistration | null>(null);
@@ -52,72 +56,50 @@ const useMp3 = (transcriptId: string, waiting?: boolean): Mp3Response => {
   }, [navigator.serviceWorker, !serviceWorker, accessTokenInfo]);
 
   useEffect(() => {
-    if (!transcriptId || !api || later) return;
+    if (!transcriptId || later || !transcript) return;
 
     let stopped = false;
     let audioElement: HTMLAudioElement | null = null;
     let handleCanPlay: (() => void) | null = null;
     let handleError: (() => void) | null = null;
 
-    setTranscriptMetadataLoading(true);
     setAudioLoading(true);
 
-    // First fetch transcript info to check if audio is deleted
-    api
-      .v1TranscriptGet({ transcriptId })
-      .then((transcript) => {
-        if (stopped) {
-          return;
-        }
+    const deleted = transcript.audio_deleted || false;
+    setAudioDeleted(deleted);
 
-        const deleted = transcript.audio_deleted || false;
-        setAudioDeleted(deleted);
-        setTranscriptMetadataLoadingError(null);
+    if (deleted) {
+      // Audio is deleted, don't attempt to load it
+      setMedia(null);
+      setAudioLoadingError(null);
+      setAudioLoading(false);
+      return;
+    }
 
-        if (deleted) {
-          // Audio is deleted, don't attempt to load it
-          setMedia(null);
-          setAudioLoadingError(null);
-          setAudioLoading(false);
-          return;
-        }
+    // Audio is not deleted, proceed to load it
+    audioElement = document.createElement("audio");
+    audioElement.src = `${api_url}/v1/transcripts/${transcriptId}/audio/mp3`;
+    audioElement.crossOrigin = "anonymous";
+    audioElement.preload = "auto";
 
-        // Audio is not deleted, proceed to load it
-        audioElement = document.createElement("audio");
-        audioElement.src = `${api_url}/v1/transcripts/${transcriptId}/audio/mp3`;
-        audioElement.crossOrigin = "anonymous";
-        audioElement.preload = "auto";
+    handleCanPlay = () => {
+      if (stopped) return;
+      setAudioLoading(false);
+      setAudioLoadingError(null);
+    };
 
-        handleCanPlay = () => {
-          if (stopped) return;
-          setAudioLoading(false);
-          setAudioLoadingError(null);
-        };
+    handleError = () => {
+      if (stopped) return;
+      setAudioLoading(false);
+      setAudioLoadingError("Failed to load audio");
+    };
 
-        handleError = () => {
-          if (stopped) return;
-          setAudioLoading(false);
-          setAudioLoadingError("Failed to load audio");
-        };
+    audioElement.addEventListener("canplay", handleCanPlay);
+    audioElement.addEventListener("error", handleError);
 
-        audioElement.addEventListener("canplay", handleCanPlay);
-        audioElement.addEventListener("error", handleError);
-
-        if (!stopped) {
-          setMedia(audioElement);
-        }
-      })
-      .catch((error) => {
-        if (stopped) return;
-        console.error("Failed to fetch transcript:", error);
-        setAudioDeleted(null);
-        setTranscriptMetadataLoadingError(error.message);
-        setAudioLoading(false);
-      })
-      .finally(() => {
-        if (stopped) return;
-        setTranscriptMetadataLoading(false);
-      });
+    if (!stopped) {
+      setMedia(audioElement);
+    }
 
     return () => {
       stopped = true;
@@ -128,14 +110,18 @@ const useMp3 = (transcriptId: string, waiting?: boolean): Mp3Response => {
         if (handleError) audioElement.removeEventListener("error", handleError);
       }
     };
-  }, [transcriptId, api, later, api_url]);
+  }, [transcriptId, transcript, later, api_url]);
 
   const getNow = () => {
     setLater(false);
   };
 
   const loading = audioLoading || transcriptMetadataLoading;
-  const error = audioLoadingError || transcriptMetadataLoadingError;
+  const error =
+    audioLoadingError ||
+    (transcriptError
+      ? (transcriptError as any).message || String(transcriptError)
+      : null);
 
   return { media, loading, error, getNow, audioDeleted };
 };

www/app/(app)/transcripts/useParticipants.ts

@@ -1,8 +1,6 @@
-import { useEffect, useState } from "react";
-import { Participant } from "../../api";
-import { useError } from "../../(errors)/errorContext";
-import useApi from "../../lib/useApi";
-import { shouldShowError } from "../../lib/errorUtils";
+import type { components } from "../../reflector-api";
+type Participant = components["schemas"]["Participant"];
+import { useTranscriptParticipants } from "../../lib/apiHooks";
 
 type ErrorParticipants = {
   error: Error;
@@ -29,46 +27,38 @@ export type UseParticipants = (
 ) & { refetch: () => void };
 
 const useParticipants = (transcriptId: string): UseParticipants => {
-  const [response, setResponse] = useState<Participant[] | null>(null);
-  const [loading, setLoading] = useState<boolean>(true);
-  const [error, setErrorState] = useState<Error | null>(null);
-  const { setError } = useError();
-  const api = useApi();
-  const [count, setCount] = useState(0);
+  const {
+    data: response,
+    isLoading: loading,
+    error,
+    refetch,
+  } = useTranscriptParticipants(transcriptId || null);
 
-  const refetch = () => {
-    if (!loading) {
-      setCount(count + 1);
-      setLoading(true);
-      setErrorState(null);
-    }
-  };
+  // Type-safe return based on state
+  if (error) {
+    return {
+      error: error as Error,
+      loading: false,
+      response: null,
+      refetch,
+    } satisfies ErrorParticipants & { refetch: () => void };
+  }
 
-  useEffect(() => {
-    if (!transcriptId || !api) return;
+  if (loading || !response) {
+    return {
+      response: response || null,
+      loading: true,
+      error: null,
+      refetch,
+    } satisfies LoadingParticipants & { refetch: () => void };
+  }
 
-    setLoading(true);
-    api
-      .v1TranscriptGetParticipants({ transcriptId })
-      .then((result) => {
-        setResponse(result);
-        setLoading(false);
-        console.debug("Participants Loaded:", result);
-      })
-      .catch((error) => {
-        const shouldShowHuman = shouldShowError(error);
-        if (shouldShowHuman) {
-          setError(error, "There was an error loading the participants");
-        } else {
-          setError(error);
-        }
-        setErrorState(error);
-        setResponse(null);
-        setLoading(false);
-      });
-  }, [transcriptId, !api, count]);
-
-  return { response, loading, error, refetch } as UseParticipants;
+  return {
+    response,
+    loading: false,
+    error: null,
+    refetch,
+  } satisfies SuccessParticipants & { refetch: () => void };
 };
 
 export default useParticipants;

www/app/(app)/transcripts/useSearchTranscripts.ts

@@ -1,123 +0,0 @@
-// this hook is not great, we want to substitute it with a proper state management solution that is also not re-invention
-
-import { useEffect, useRef, useState } from "react";
-import { SearchResult, SourceKind } from "../../api";
-import useApi from "../../lib/useApi";
-import {
-  PaginationPage,
-  paginationPageTo0Based,
-} from "../browse/_components/Pagination";
-
-interface SearchFilters {
-  roomIds: readonly string[] | null;
-  sourceKind: SourceKind | null;
-}
-
-const EMPTY_SEARCH_FILTERS: SearchFilters = {
-  roomIds: null,
-  sourceKind: null,
-};
-
-type UseSearchTranscriptsOptions = {
-  pageSize: number;
-  page: PaginationPage;
-};
-
-interface UseSearchTranscriptsReturn {
-  results: SearchResult[];
-  totalCount: number;
-  isLoading: boolean;
-  error: unknown;
-  reload: () => void;
-}
-
-function hashEffectFilters(filters: SearchFilters): string {
-  return JSON.stringify(filters);
-}
-
-export function useSearchTranscripts(
-  query: string = "",
-  filters: SearchFilters = EMPTY_SEARCH_FILTERS,
-  options: UseSearchTranscriptsOptions = {
-    pageSize: 20,
-    page: PaginationPage(1),
-  },
-): UseSearchTranscriptsReturn {
-  const { pageSize, page } = options;
-
-  const [reloadCount, setReloadCount] = useState(0);
-
-  const api = useApi();
-  const abortControllerRef = useRef<AbortController>();
-
-  const [data, setData] = useState<{ results: SearchResult[]; total: number }>({
-    results: [],
-    total: 0,
-  });
-  const [error, setError] = useState<any>();
-  const [isLoading, setIsLoading] = useState(false);
-
-  const filterHash = hashEffectFilters(filters);
-
-  useEffect(() => {
-    if (!api) {
-      setData({ results: [], total: 0 });
-      setError(undefined);
-      setIsLoading(false);
-      return;
-    }
-
-    if (abortControllerRef.current) {
-      abortControllerRef.current.abort();
-    }
-
-    const abortController = new AbortController();
-    abortControllerRef.current = abortController;
-
-    const performSearch = async () => {
-      setIsLoading(true);
-
-      try {
-        const response = await api.v1TranscriptsSearch({
-          q: query || "",
-          limit: pageSize,
-          offset: paginationPageTo0Based(page) * pageSize,
-          roomId: filters.roomIds?.[0],
-          sourceKind: filters.sourceKind || undefined,
-        });
-
-        if (abortController.signal.aborted) return;
-        setData(response);
-        setError(undefined);
-      } catch (err: unknown) {
-        if ((err as Error).name === "AbortError") {
-          return;
-        }
-        if (abortController.signal.aborted) {
-          console.error("Aborted search but error", err);
-          return;
-        }
-
-        setError(err);
-      } finally {
-        if (!abortController.signal.aborted) {
-          setIsLoading(false);
-        }
-      }
-    };
-
-    performSearch().then(() => {});
-
-    return () => {
-      abortController.abort();
-    };
-  }, [api, query, page, filterHash, pageSize, reloadCount]);
-
-  return {
-    results: data.results,
-    totalCount: data.total,
-    isLoading,
-    error,
-    reload: () => setReloadCount(reloadCount + 1),
-  };
-}

www/app/(app)/transcripts/useTopicWithWords.ts

@@ -1,9 +1,8 @@
-import { useEffect, useState } from "react";
+import type { components } from "../../reflector-api";
+import { useTranscriptTopicsWithWordsPerSpeaker } from "../../lib/apiHooks";
 
-import { GetTranscriptTopicWithWordsPerSpeaker } from "../../api";
-import { useError } from "../../(errors)/errorContext";
-import useApi from "../../lib/useApi";
-import { shouldShowError } from "../../lib/errorUtils";
+type GetTranscriptTopicWithWordsPerSpeaker =
+  components["schemas"]["GetTranscriptTopicWithWordsPerSpeaker"];
 
 type ErrorTopicWithWords = {
   error: Error;
@@ -33,47 +32,40 @@ const useTopicWithWords = (
   topicId: string | undefined,
   transcriptId: string,
 ): UseTopicWithWords => {
-  const [response, setResponse] =
-    useState<GetTranscriptTopicWithWordsPerSpeaker | null>(null);
-  const [loading, setLoading] = useState<boolean>(false);
-  const [error, setErrorState] = useState<Error | null>(null);
-  const { setError } = useError();
-  const api = useApi();
+  const {
+    data: response,
+    isLoading: loading,
+    error,
+    refetch,
+  } = useTranscriptTopicsWithWordsPerSpeaker(
+    transcriptId || null,
+    topicId || null,
+  );
 
-  const [count, setCount] = useState(0);
+  if (error) {
+    return {
+      error: error as Error,
+      loading: false,
+      response: null,
+      refetch,
+    } satisfies ErrorTopicWithWords & { refetch: () => void };
+  }
 
-  const refetch = () => {
-    if (!loading) {
-      setCount(count + 1);
-      setLoading(true);
-      setErrorState(null);
-    }
-  };
+  if (loading || !response) {
+    return {
+      response: response || null,
+      loading: true,
+      error: false,
+      refetch,
+    } satisfies LoadingTopicWithWords & { refetch: () => void };
+  }
 
-  useEffect(() => {
-    if (!transcriptId || !topicId || !api) return;
-
-    setLoading(true);
-
-    api
-      .v1TranscriptGetTopicsWithWordsPerSpeaker({ transcriptId, topicId })
-      .then((result) => {
-        setResponse(result);
-        setLoading(false);
-        console.debug("Topics with words Loaded:", result);
-      })
-      .catch((error) => {
-        const shouldShowHuman = shouldShowError(error);
-        if (shouldShowHuman) {
-          setError(error, "There was an error loading the topics with words");
-        } else {
-          setError(error);
-        }
-        setErrorState(error);
-      });
-  }, [transcriptId, !api, topicId, count]);
-
-  return { response, loading, error, refetch } as UseTopicWithWords;
+  return {
+    response,
+    loading: false,
+    error: null,
+    refetch,
+  } satisfies SuccessTopicWithWords & { refetch: () => void };
 };
 
 export default useTopicWithWords;

www/app/(app)/transcripts/useTopics.ts

@@ -1,10 +1,7 @@
-import { useEffect, useState } from "react";
+import { useTranscriptTopics } from "../../lib/apiHooks";
+import type { components } from "../../reflector-api";
 
-import { useError } from "../../(errors)/errorContext";
-import { Topic } from "./webSocketTypes";
-import useApi from "../../lib/useApi";
-import { shouldShowError } from "../../lib/errorUtils";
-import { GetTranscriptTopic } from "../../api";
+type GetTranscriptTopic = components["schemas"]["GetTranscriptTopic"];
 
 type TranscriptTopics = {
   topics: GetTranscriptTopic[] | null;
@@ -13,34 +10,13 @@ type TranscriptTopics = {
 };
 
 const useTopics = (id: string): TranscriptTopics => {
-  const [topics, setTopics] = useState<Topic[] | null>(null);
-  const [loading, setLoading] = useState<boolean>(true);
-  const [error, setErrorState] = useState<Error | null>(null);
-  const { setError } = useError();
-  const api = useApi();
-  useEffect(() => {
-    if (!id || !api) return;
+  const { data: topics, isLoading: loading, error } = useTranscriptTopics(id);
 
-    setLoading(true);
-    api
-      .v1TranscriptGetTopics({ transcriptId: id })
-      .then((result) => {
-        setTopics(result);
-        setLoading(false);
-        console.debug("Transcript topics loaded:", result);
-      })
-      .catch((err) => {
-        setErrorState(err);
-        const shouldShowHuman = shouldShowError(err);
-        if (shouldShowHuman) {
-          setError(err, "There was an error loading the topics");
-        } else {
-          setError(err);
-        }
-      });
-  }, [id, !api]);
-
-  return { topics, loading, error };
+  return {
+    topics: topics || null,
+    loading,
+    error: error as Error | null,
+  };
 };
 
 export default useTopics;

www/app/(app)/transcripts/useTranscript.ts

@@ -1,8 +1,7 @@
-import { useEffect, useState } from "react";
-import { GetTranscript } from "../../api";
-import { useError } from "../../(errors)/errorContext";
-import { shouldShowError } from "../../lib/errorUtils";
-import useApi from "../../lib/useApi";
+import type { components } from "../../reflector-api";
+import { useTranscriptGet } from "../../lib/apiHooks";
+
+type GetTranscript = components["schemas"]["GetTranscript"];
 
 type ErrorTranscript = {
   error: Error;
@@ -28,43 +27,43 @@ type SuccessTranscript = {
 const useTranscript = (
   id: string | null,
 ): ErrorTranscript | LoadingTranscript | SuccessTranscript => {
-  const [response, setResponse] = useState<GetTranscript | null>(null);
-  const [loading, setLoading] = useState<boolean>(true);
-  const [error, setErrorState] = useState<Error | null>(null);
-  const [reload, setReload] = useState(0);
-  const { setError } = useError();
-  const api = useApi();
-  const reloadHandler = () => setReload((prev) => prev + 1);
+  const { data, isLoading, error, refetch } = useTranscriptGet(id);
 
-  useEffect(() => {
-    if (!id || !api) return;
+  // Map to the expected return format
+  if (isLoading) {
+    return {
+      response: null,
+      loading: true,
+      error: false,
+      reload: refetch,
+    };
+  }
 
-    if (!response) {
-      setLoading(true);
-    }
+  if (error) {
+    return {
+      error: error as Error,
+      loading: false,
+      response: null,
+      reload: refetch,
+    };
+  }
 
-    api
-      .v1TranscriptGet({ transcriptId: id })
-      .then((result) => {
-        setResponse(result);
-        setLoading(false);
-        console.debug("Transcript Loaded:", result);
-      })
-      .catch((error) => {
-        const shouldShowHuman = shouldShowError(error);
-        if (shouldShowHuman) {
-          setError(error, "There was an error loading the transcript");
-        } else {
-          setError(error);
-        }
-        setErrorState(error);
-      });
-  }, [id, !api, reload]);
+  // Check if data is undefined or null
+  if (!data) {
+    return {
+      response: null,
+      loading: true,
+      error: false,
+      reload: refetch,
+    };
+  }
 
-  return { response, loading, error, reload: reloadHandler } as
-    | ErrorTranscript
-    | LoadingTranscript
-    | SuccessTranscript;
+  return {
+    response: data,
+    loading: false,
+    error: null,
+    reload: refetch,
+  };
 };
 
 export default useTranscript;

www/app/(app)/transcripts/useWaveform.ts

@@ -1,8 +1,7 @@
-import { useEffect, useState } from "react";
-import { AudioWaveform } from "../../api";
-import { useError } from "../../(errors)/errorContext";
-import useApi from "../../lib/useApi";
-import { shouldShowError } from "../../lib/errorUtils";
+import type { components } from "../../reflector-api";
+import { useTranscriptWaveform } from "../../lib/apiHooks";
+
+type AudioWaveform = components["schemas"]["AudioWaveform"];
 
 type AudioWaveFormResponse = {
   waveform: AudioWaveform | null;
@@ -11,35 +10,17 @@ type AudioWaveFormResponse = {
 };
 
 const useWaveform = (id: string, skip: boolean): AudioWaveFormResponse => {
-  const [waveform, setWaveform] = useState<AudioWaveform | null>(null);
-  const [loading, setLoading] = useState<boolean>(false);
-  const [error, setErrorState] = useState<Error | null>(null);
-  const { setError } = useError();
-  const api = useApi();
+  const {
+    data: waveform,
+    isLoading: loading,
+    error,
+  } = useTranscriptWaveform(skip ? null : id);
 
-  useEffect(() => {
-    if (!id || !api || skip) {
-      setLoading(false);
-      setErrorState(null);
-      setWaveform(null);
-      return;
-    }
-    setLoading(true);
-    setErrorState(null);
-    api
-      .v1TranscriptGetAudioWaveform({ transcriptId: id })
-      .then((result) => {
-        setWaveform(result);
-        setLoading(false);
-        console.debug("Transcript waveform loaded:", result);
-      })
-      .catch((err) => {
-        setErrorState(err);
-        setLoading(false);
-      });
-  }, [id, api, skip]);
-
-  return { waveform, loading, error };
+  return {
+    waveform: waveform || null,
+    loading,
+    error: error as Error | null,
+  };
 };
 
 export default useWaveform;

www/app/(app)/transcripts/useWebRTC.ts

@@ -1,8 +1,9 @@
 import { useEffect, useState } from "react";
 import Peer from "simple-peer";
 import { useError } from "../../(errors)/errorContext";
-import useApi from "../../lib/useApi";
-import { RtcOffer } from "../../api";
+import { useTranscriptWebRTC } from "../../lib/apiHooks";
+import type { components } from "../../reflector-api";
+type RtcOffer = components["schemas"]["RtcOffer"];
 
 const useWebRTC = (
   stream: MediaStream | null,
@@ -10,10 +11,10 @@ const useWebRTC = (
 ): Peer => {
   const [peer, setPeer] = useState<Peer | null>(null);
   const { setError } = useError();
-  const api = useApi();
+  const { mutateAsync: mutateWebRtcTranscriptAsync } = useTranscriptWebRTC();
 
   useEffect(() => {
-    if (!stream || !transcriptId || !api) {
+    if (!stream || !transcriptId) {
       return;
     }
 
@@ -24,7 +25,7 @@ const useWebRTC = (
     try {
       p = new Peer({ initiator: true, stream: stream });
     } catch (error) {
-      setError(error, "Error creating WebRTC");
+      setError(error as Error, "Error creating WebRTC");
       return;
     }
 
@@ -32,26 +33,31 @@ const useWebRTC = (
       setError(new Error(`WebRTC error: ${err}`));
     });
 
-    p.on("signal", (data: any) => {
-      if (!api) return;
+    p.on("signal", async (data: any) => {
       if ("sdp" in data) {
         const rtcOffer: RtcOffer = {
           sdp: data.sdp,
           type: data.type,
         };
 
-        api
-          .v1TranscriptRecordWebrtc({ transcriptId, requestBody: rtcOffer })
-          .then((answer) => {
-            try {
-              p.signal(answer);
-            } catch (error) {
-              setError(error);
-            }
-          })
-          .catch((error) => {
-            setError(error, "Error loading WebRTCOffer");
+        try {
+          const answer = await mutateWebRtcTranscriptAsync({
+            params: {
+              path: {
+                transcript_id: transcriptId,
+              },
+            },
+            body: rtcOffer,
           });
+
+          try {
+            p.signal(answer);
+          } catch (error) {
+            setError(error as Error);
+          }
+        } catch (error) {
+          setError(error as Error, "Error loading WebRTCOffer");
+        }
       }
     });
 
@@ -63,7 +69,7 @@ const useWebRTC = (
     return () => {
       p.destroy();
     };
-  }, [stream, transcriptId, !api]);
+  }, [stream, transcriptId, mutateWebRtcTranscriptAsync]);
 
   return peer;
 };

www/app/(app)/transcripts/useWebSockets.ts

@@ -2,8 +2,12 @@ import { useContext, useEffect, useState } from "react";
 import { Topic, FinalSummary, Status } from "./webSocketTypes";
 import { useError } from "../../(errors)/errorContext";
 import { DomainContext } from "../../domainContext";
-import { AudioWaveform, GetTranscriptSegmentTopic } from "../../api";
-import useApi from "../../lib/useApi";
+import type { components } from "../../reflector-api";
+type AudioWaveform = components["schemas"]["AudioWaveform"];
+type GetTranscriptSegmentTopic =
+  components["schemas"]["GetTranscriptSegmentTopic"];
+import { useQueryClient } from "@tanstack/react-query";
+import { $api } from "../../lib/apiClient";
 
 export type UseWebSockets = {
   transcriptTextLive: string;
@@ -33,8 +37,8 @@ export const useWebSockets = (transcriptId: string | null): UseWebSockets => {
   const [status, setStatus] = useState<Status>({ value: "" });
   const { setError } = useError();
 
-  const { websocket_url } = useContext(DomainContext);
-  const api = useApi();
+  const { websocket_url: websocketUrl } = useContext(DomainContext);
+  const queryClient = useQueryClient();
 
   const [accumulatedText, setAccumulatedText] = useState<string>("");
 
@@ -105,6 +109,13 @@ export const useWebSockets = (transcriptId: string | null): UseWebSockets => {
             title: "Topic 1: Introduction to Quantum Mechanics",
             transcript:
               "A brief overview of quantum mechanics and its principles.",
+            segments: [
+              {
+                speaker: 1,
+                start: 0,
+                text: "This is the transcription of an example title",
+              },
+            ],
           },
           {
             id: "2",
@@ -315,11 +326,9 @@ export const useWebSockets = (transcriptId: string | null): UseWebSockets => {
       }
     };
 
-    if (!transcriptId || !api) return;
+    if (!transcriptId) return;
 
-    api?.v1TranscriptGetWebsocketEvents({ transcriptId }).then((result) => {});
-
-    const url = `${websocket_url}/v1/transcripts/${transcriptId}/events`;
+    const url = `${websocketUrl}/v1/transcripts/${transcriptId}/events`;
     let ws = new WebSocket(url);
 
     ws.onopen = () => {
@@ -361,6 +370,16 @@ export const useWebSockets = (transcriptId: string | null): UseWebSockets => {
               return [...prevTopics, topic];
             });
             console.debug("TOPIC event:", message.data);
+            // Invalidate topics query to sync with WebSocket data
+            queryClient.invalidateQueries({
+              queryKey: $api.queryOptions(
+                "get",
+                "/v1/transcripts/{transcript_id}/topics",
+                {
+                  params: { path: { transcript_id: transcriptId } },
+                },
+              ).queryKey,
+            });
             break;
 
           case "FINAL_SHORT_SUMMARY":
@@ -370,6 +389,16 @@ export const useWebSockets = (transcriptId: string | null): UseWebSockets => {
           case "FINAL_LONG_SUMMARY":
             if (message.data) {
               setFinalSummary(message.data);
+              // Invalidate transcript query to sync summary
+              queryClient.invalidateQueries({
+                queryKey: $api.queryOptions(
+                  "get",
+                  "/v1/transcripts/{transcript_id}",
+                  {
+                    params: { path: { transcript_id: transcriptId } },
+                  },
+                ).queryKey,
+              });
             }
             break;
 
@@ -377,6 +406,16 @@ export const useWebSockets = (transcriptId: string | null): UseWebSockets => {
             console.debug("FINAL_TITLE event:", message.data);
             if (message.data) {
               setTitle(message.data.title);
+              // Invalidate transcript query to sync title
+              queryClient.invalidateQueries({
+                queryKey: $api.queryOptions(
+                  "get",
+                  "/v1/transcripts/{transcript_id}",
+                  {
+                    params: { path: { transcript_id: transcriptId } },
+                  },
+                ).queryKey,
+              });
             }
             break;
 
@@ -434,6 +473,11 @@ export const useWebSockets = (transcriptId: string | null): UseWebSockets => {
           break;
         case 1001: // Navigate away
           break;
+        case 1006: // Closed by client Chrome
+          console.warn(
+            "WebSocket closed by client, likely duplicated connection in react dev mode",
+          );
+          break;
         default:
           setError(
             new Error(`WebSocket closed unexpectedly with code: ${event.code}`),
@@ -450,7 +494,7 @@ export const useWebSockets = (transcriptId: string | null): UseWebSockets => {
     return () => {
       ws.close();
     };
-  }, [transcriptId, !api]);
+  }, [transcriptId, websocketUrl]);
 
   return {
     transcriptTextLive,

www/app/(app)/transcripts/webSocketTypes.ts

@@ -1,4 +1,6 @@
-import { GetTranscriptTopic } from "../../api";
+import type { components } from "../../reflector-api";
+
+type GetTranscriptTopic = components["schemas"]["GetTranscriptTopic"];
 
 export type Topic = GetTranscriptTopic;
 

www/app/(auth)/userInfo.tsx

@@ -1,18 +1,21 @@
 "use client";
-import { signOut, signIn } from "next-auth/react";
-import useSessionStatus from "../lib/useSessionStatus";
+
 import { Spinner, Link } from "@chakra-ui/react";
+import { useAuth } from "../lib/AuthProvider";
 
 export default function UserInfo() {
-  const { isLoading, isAuthenticated } = useSessionStatus();
-
+  const auth = useAuth();
+  const status = auth.status;
+  const isLoading = status === "loading";
+  const isAuthenticated = status === "authenticated";
+  const isRefreshing = status === "refreshing";
   return isLoading ? (
     <Spinner size="xs" className="mx-3" />
-  ) : !isAuthenticated ? (
+  ) : !isAuthenticated && !isRefreshing ? (
     <Link
       href="/"
       className="font-light px-2"
-      onClick={() => signIn("authentik")}
+      onClick={() => auth.signIn("authentik")}
     >
       Log in
     </Link>
@@ -20,7 +23,7 @@ export default function UserInfo() {
     <Link
       href="#"
       className="font-light px-2"
-      onClick={() => signOut({ callbackUrl: "/" })}
+      onClick={() => auth.signOut({ callbackUrl: "/" })}
     >
       Log out
     </Link>

www/app/[roomName]/page.tsx

@@ -21,11 +21,13 @@ import { toaster } from "../components/ui/toaster";
 import useRoomMeeting from "./useRoomMeeting";
 import { useRouter } from "next/navigation";
 import { notFound } from "next/navigation";
-import useSessionStatus from "../lib/useSessionStatus";
 import { useRecordingConsent } from "../recordingConsentContext";
-import useApi from "../lib/useApi";
-import { Meeting } from "../api";
+import { useMeetingAudioConsent } from "../lib/apiHooks";
+import type { components } from "../reflector-api";
+
+type Meeting = components["schemas"]["Meeting"];
 import { FaBars } from "react-icons/fa6";
+import { useAuth } from "../lib/AuthProvider";
 
 export type RoomDetails = {
   params: {
@@ -76,31 +78,30 @@ const useConsentDialog = (
   wherebyRef: RefObject<HTMLElement> /*accessibility*/,
 ) => {
   const { state: consentState, touch, hasConsent } = useRecordingConsent();
-  const [consentLoading, setConsentLoading] = useState(false);
   // toast would open duplicates, even with using "id=" prop
   const [modalOpen, setModalOpen] = useState(false);
-  const api = useApi();
+  const audioConsentMutation = useMeetingAudioConsent();
 
   const handleConsent = useCallback(
     async (meetingId: string, given: boolean) => {
-      if (!api) return;
-
-      setConsentLoading(true);
-
       try {
-        await api.v1MeetingAudioConsent({
-          meetingId,
-          requestBody: { consent_given: given },
+        await audioConsentMutation.mutateAsync({
+          params: {
+            path: {
+              meeting_id: meetingId,
+            },
+          },
+          body: {
+            consent_given: given,
+          },
         });
 
         touch(meetingId);
       } catch (error) {
         console.error("Error submitting consent:", error);
-      } finally {
-        setConsentLoading(false);
       }
     },
-    [api, touch],
+    [audioConsentMutation, touch],
   );
 
   const showConsentModal = useCallback(() => {
@@ -194,7 +195,12 @@ const useConsentDialog = (
     return cleanup;
   }, [meetingId, handleConsent, wherebyRef, modalOpen]);
 
-  return { showConsentModal, consentState, hasConsent, consentLoading };
+  return {
+    showConsentModal,
+    consentState,
+    hasConsent,
+    consentLoading: audioConsentMutation.isPending,
+  };
 };
 
 function ConsentDialogButton({
@@ -254,7 +260,9 @@ export default function Room(details: RoomDetails) {
   const roomName = details.params.roomName;
   const meeting = useRoomMeeting(roomName);
   const router = useRouter();
-  const { isLoading, isAuthenticated } = useSessionStatus();
+  const status = useAuth().status;
+  const isAuthenticated = status === "authenticated";
+  const isLoading = status === "loading" || meeting.loading;
 
   const roomUrl = meeting?.response?.host_room_url
     ? meeting?.response?.host_room_url

www/app/[roomName]/useRoomMeeting.tsx

@@ -1,8 +1,10 @@
 import { useEffect, useState } from "react";
 import { useError } from "../(errors)/errorContext";
-import { Meeting } from "../api";
+import type { components } from "../reflector-api";
 import { shouldShowError } from "../lib/errorUtils";
-import useApi from "../lib/useApi";
+
+type Meeting = components["schemas"]["Meeting"];
+import { useRoomsCreateMeeting } from "../lib/apiHooks";
 import { notFound } from "next/navigation";
 
 type ErrorMeeting = {
@@ -30,27 +32,25 @@ const useRoomMeeting = (
   roomName: string | null | undefined,
 ): ErrorMeeting | LoadingMeeting | SuccessMeeting => {
   const [response, setResponse] = useState<Meeting | null>(null);
-  const [loading, setLoading] = useState<boolean>(true);
-  const [error, setErrorState] = useState<Error | null>(null);
   const [reload, setReload] = useState(0);
   const { setError } = useError();
-  const api = useApi();
+  const createMeetingMutation = useRoomsCreateMeeting();
   const reloadHandler = () => setReload((prev) => prev + 1);
 
   useEffect(() => {
-    if (!roomName || !api) return;
+    if (!roomName) return;
 
-    if (!response) {
-      setLoading(true);
-    }
-
-    api
-      .v1RoomsCreateMeeting({ roomName })
-      .then((result) => {
+    const createMeeting = async () => {
+      try {
+        const result = await createMeetingMutation.mutateAsync({
+          params: {
+            path: {
+              room_name: roomName,
+            },
+          },
+        });
         setResponse(result);
-        setLoading(false);
-      })
-      .catch((error) => {
+      } catch (error: any) {
         const shouldShowHuman = shouldShowError(error);
         if (shouldShowHuman && error.status !== 404) {
           setError(
@@ -60,9 +60,14 @@ const useRoomMeeting = (
         } else {
           setError(error);
         }
-        setErrorState(error);
-      });
-  }, [roomName, !api, reload]);
+      }
+    };
+
+    createMeeting();
+  }, [roomName, reload]);
+
+  const loading = createMeetingMutation.isPending && !response;
+  const error = createMeetingMutation.error as Error | null;
 
   return { response, loading, error, reload: reloadHandler } as
     | ErrorMeeting

www/app/api/OpenApi.ts

@@ -1,37 +0,0 @@
-import type { BaseHttpRequest } from "./core/BaseHttpRequest";
-import type { OpenAPIConfig } from "./core/OpenAPI";
-import { Interceptors } from "./core/OpenAPI";
-import { AxiosHttpRequest } from "./core/AxiosHttpRequest";
-
-import { DefaultService } from "./services.gen";
-
-type HttpRequestConstructor = new (config: OpenAPIConfig) => BaseHttpRequest;
-
-export class OpenApi {
-  public readonly default: DefaultService;
-
-  public readonly request: BaseHttpRequest;
-
-  constructor(
-    config?: Partial<OpenAPIConfig>,
-    HttpRequest: HttpRequestConstructor = AxiosHttpRequest,
-  ) {
-    this.request = new HttpRequest({
-      BASE: config?.BASE ?? "",
-      VERSION: config?.VERSION ?? "0.1.0",
-      WITH_CREDENTIALS: config?.WITH_CREDENTIALS ?? false,
-      CREDENTIALS: config?.CREDENTIALS ?? "include",
-      TOKEN: config?.TOKEN,
-      USERNAME: config?.USERNAME,
-      PASSWORD: config?.PASSWORD,
-      HEADERS: config?.HEADERS,
-      ENCODE_PATH: config?.ENCODE_PATH,
-      interceptors: {
-        request: config?.interceptors?.request ?? new Interceptors(),
-        response: config?.interceptors?.response ?? new Interceptors(),
-      },
-    });
-
-    this.default = new DefaultService(this.request);
-  }
-}

www/app/api/auth/[...nextauth]/route.ts

@@ -1,8 +1,5 @@
-// NextAuth route handler for Authentik
-// Refresh rotation has been taken from https://next-auth.js.org/v3/tutorials/refresh-token-rotation even if we are using 4.x
-
 import NextAuth from "next-auth";
-import { authOptions } from "../../../lib/auth";
+import { authOptions } from "../../../lib/authBackend";
 
 const handler = NextAuth(authOptions);
 

www/app/api/core/ApiError.ts

@@ -1,25 +0,0 @@
-import type { ApiRequestOptions } from "./ApiRequestOptions";
-import type { ApiResult } from "./ApiResult";
-
-export class ApiError extends Error {
-  public readonly url: string;
-  public readonly status: number;
-  public readonly statusText: string;
-  public readonly body: unknown;
-  public readonly request: ApiRequestOptions;
-
-  constructor(
-    request: ApiRequestOptions,
-    response: ApiResult,
-    message: string,
-  ) {
-    super(message);
-
-    this.name = "ApiError";
-    this.url = response.url;
-    this.status = response.status;
-    this.statusText = response.statusText;
-    this.body = response.body;
-    this.request = request;
-  }
-}

www/app/api/core/ApiRequestOptions.ts

@@ -1,21 +0,0 @@
-export type ApiRequestOptions<T = unknown> = {
-  readonly method:
-    | "GET"
-    | "PUT"
-    | "POST"
-    | "DELETE"
-    | "OPTIONS"
-    | "HEAD"
-    | "PATCH";
-  readonly url: string;
-  readonly path?: Record<string, unknown>;
-  readonly cookies?: Record<string, unknown>;
-  readonly headers?: Record<string, unknown>;
-  readonly query?: Record<string, unknown>;
-  readonly formData?: Record<string, unknown>;
-  readonly body?: any;
-  readonly mediaType?: string;
-  readonly responseHeader?: string;
-  readonly responseTransformer?: (data: unknown) => Promise<T>;
-  readonly errors?: Record<number | string, string>;
-};

www/app/api/core/ApiResult.ts

@@ -1,7 +0,0 @@
-export type ApiResult<TData = any> = {
-  readonly body: TData;
-  readonly ok: boolean;
-  readonly status: number;
-  readonly statusText: string;
-  readonly url: string;
-};