Teams Meeting Transcription App

Single-container application for transcribing Microsoft Teams recordings using FFmpeg.js and OpenAI Whisper API.

Features

• Upload Teams recordings (MP4, AVI, MOV, etc.)
• Client-side audio extraction using FFmpeg.js
• High-accuracy transcription via OpenAI Whisper API
• Single Docker container - easy deployment
• Clean web interface with drag-and-drop upload
• Copy/download results as text files

Quick Start

1. Set OpenAI API Key

export OPENAI_API_KEY="your_openai_api_key_here"

2. Build & Run with Docker

# Build image
docker build -t transcription-app .

# Run container
docker run -d \
  -p 3000:3000 \
  -e OPENAI_API_KEY=$OPENAI_API_KEY \
  -v $(pwd)/uploads:/app/uploads \
  --name transcription \
  transcription-app

3. Access Application

Open http://localhost:3000 in your browser

QNAP Container Station Deployment

1. Upload project folder to QNAP
2. Container Station → Create → Build Image
3. Set environment variable: OPENAI_API_KEY
4. Map volume: /share/transcriptions → /app/uploads
5. Expose port: 3000

Usage

1. Upload Teams recording file (video or audio)
2. Wait for FFmpeg to extract audio (if video)
3. Transcription happens automatically via Whisper API
4. Copy or download the result text

File Size Limits

• Max upload: 100MB
• Recommended: Use MP3/M4A for faster processing
• Cost: ~$0.36 per hour of audio

Supported Formats

Input: MP4, AVI, MOV, WMV, MKV, MP3, WAV, M4A, AAC Output: Compressed MP3 (128kbps) → Whisper API

Architecture

Browser → FFmpeg.js (audio extraction) → Node.js API → OpenAI Whisper

Environment Variables

• OPENAI_API_KEY - Required for transcription
• PORT - Server port (default: 3000)

Development

# Install dependencies
npm install

# Run locally
npm start

# Access at http://localhost:3000

Troubleshooting

FFmpeg loading issues: Refresh browser, check console Transcription errors: Verify OpenAI API key and credits Large files: Consider pre-converting to MP3 for faster processing QNAP deployment: Ensure Container Station has sufficient RAM (4GB+)