📝 Meet Summarizer

A meeting recording, transcription, and summarization system. Record your screen/audio with FFmpeg, transcribe with Whisper, and summarize with OpenAI.

🚀 Quick Start

1. Installation

# Install dependencies
pip install -r requirements.txt

# Set up environment variables
cp .env.example .env
nano .env  # Add your OpenAI API key

2. FFmpeg Installation (macOS)

# Install FFmpeg with Homebrew
brew install ffmpeg

# Check screen recording permissions
# System Settings > Security & Privacy > Screen Recording

3. OpenAI API Key

1. Get your API key from OpenAI Platform
2. Add it to your .env file:

OPENAI_API_KEY=sk-your-api-key-here

📖 Usage

Command Line Interface

# Help
python main.py --help

# Start screen recording
python main.py record

# Timed recording (60 seconds)
python main.py record --duration 60

# Extract audio from video
python main.py extract video.mp4

# Transcribe audio file
python main.py transcribe audio.wav --language en

# Summarize transcript
python main.py summarize transcript.txt --type comprehensive

# Full pipeline (Video → Audio → Transcript → Summary)
python main.py process video.mp4 --language en --summary-type brief

Manual Usage

# 1. Screen recording
python screen_recorder.py

# 2. Extract audio
python audio_extractor.py video.mp4

# 3. Transcription
python transcriber.py audio.wav en

# 4. Summarization
python summarizer.py transcript.txt en comprehensive

📁 Folder Structure

meet-summarizer/
├── recordings/     # Video recordings
├── audio/          # Audio files
├── transcripts/    # Transcription files
└── summaries/      # AI summaries

⚙️ Configuration

Whisper Models

• tiny: Fastest, lowest quality
• base: Balanced (default)
• small: Good quality
• medium: High quality
• large: Best quality, slowest

Summary Types

• comprehensive: Detailed summary (default)
• brief: Short summary (200 words)
• action_items: Action items only

Supported Languages

• en: English
• tr: Turkish
• de: German
• fr: French
• es: Spanish
• and more...

🎯 Features

✅ Screen Recording

• Uses macOS AVFoundation
• Records both screen and audio
• Duration limit
• Stop with Ctrl+C

✅ Audio Processing

• High-quality audio extraction with FFmpeg
• Multiple format support
• Mono audio (optimal for Whisper)

✅ Transcription

• Uses OpenAI Whisper
• Word-level timestamps
• SRT subtitle output
• Multi-language support

✅ AI Summarization

• Smart summaries with GPT-4
• Multiple summary types
• Token usage tracking
• Sentiment analysis

🔧 Troubleshooting

FFmpeg Errors

# List audio devices on macOS
python main.py devices

# Check permissions
# System Settings > Security & Privacy > Screen Recording

Whisper is Slow

# Use a smaller model
python main.py transcribe audio.wav --model tiny

OpenAI API Error

# Check your API key
echo $OPENAI_API_KEY

# Check your usage/quota
# https://platform.openai.com/usage

📊 System Status

# File counts and sizes
python main.py status

🚨 Notes

1. Privacy: Recordings are stored locally only
2. API Cost: OpenAI API is a paid service
3. Performance: Large files may take time to process
4. Disk Space: Video files can be large

📄 License

MIT License - See LICENSE for details.

🤝 Contributing

1. Fork this repo
2. Create a feature branch
3. Commit your changes
4. Open a pull request

📞 Support

For issues, please use GitHub Issues.