Call Auditing Application

A comprehensive audio call auditing system that provides speaker diarization and transcription capabilities using two different approaches:

1. OpenAI Pipeline: Uses OpenAI's Whisper model for both transcription and diarization
2. PyAnnote + OpenAI Pipeline: Uses PyAnnote for speaker diarization and OpenAI Whisper for transcription

🚀 Features

• Speaker Diarization: Identify and separate different speakers in audio recordings
• High-Quality Transcription: Convert speech to text with timestamp accuracy
• Two Implementation Approaches: Choose between OpenAI-only or hybrid PyAnnote+OpenAI solutions
• Audio Preprocessing: Noise reduction, VAD (Voice Activity Detection), and audio cleanup
• Batch Processing: Handle multiple audio files efficiently
• Detailed Analytics: Generate comprehensive meeting summaries and insights
• Export Capabilities: Multiple output formats (JSON, CSV, TXT)

📁 Project Structure

openai-diarize/
├── openai/                          # OpenAI-based pipeline
│   ├── OpenAI_Diarized_Pipeline.ipynb
│   ├── openai_pipeline.py          # Converted Python script
│   └── pyproject.toml
├── pyannote/                        # PyAnnote + OpenAI pipeline  
│   ├── Diarize.ipynb
│   ├── transcribe.ipynb
│   ├── diarize.py                   # Converted Python script
│   ├── transcribe.py                # Converted Python script
│   └── pyproject.toml
├── requirements.txt                 # Combined dependencies
├── .env.example                     # Environment variables template
├── .gitignore
└── README.md

🛠️ Installation

Prerequisites

• Python 3.11 or higher
• FFmpeg (for audio processing)
• OpenAI API key
• Hugging Face API key (for PyAnnote approach)

Setup

1. Clone the repository

   git clone https://github.com/your-username/openai-diarize.git
   cd openai-diarize

2. Install dependencies

   Option A: Using pip

   pip install -r requirements.txt

   Option B: Using Poetry (recommended)

   # For OpenAI pipeline
   cd openai && poetry install
   
   # For PyAnnote pipeline  
   cd ../pyannote && poetry install

3. Set up environment variables

   cp .env.example .env

   Edit .env and add your API keys:

   OPENAI_API_KEY=your_openai_api_key_here
   HUGGINGFACE_API_KEY=your_huggingface_api_key_here
   

4. Install FFmpeg

   • Windows: Download from https://ffmpeg.org/download.html
   • macOS: brew install ffmpeg
   • Ubuntu/Debian: sudo apt install ffmpeg

🔧 Usage

OpenAI Pipeline

The OpenAI pipeline uses OpenAI's Whisper model for both diarization and transcription:

Using Python Script:

cd openai
python openai_pipeline.py --input path/to/your/audio.mp4 --output results/

Using Jupyter Notebook:

cd openai
jupyter notebook OpenAI_Diarized_Pipeline.ipynb

PyAnnote + OpenAI Pipeline

This approach uses PyAnnote for more accurate speaker diarization and OpenAI for transcription:

Step 1: Diarization

cd pyannote
python diarize.py --input path/to/your/audio.mp4

Step 2: Transcription

python transcribe.py

Using Jupyter Notebooks:

cd pyannote
jupyter notebook Diarize.ipynb
# Then run transcribe.ipynb

📊 Output Formats

The application generates multiple output formats:

• JSON: Detailed transcription with timestamps and speaker information
• CSV: Structured data for analysis
• TXT: Clean, readable meeting summary
• Analytics: Speaker insights, conversation metrics, and summaries

🔍 Key Features Comparison

Feature	OpenAI Pipeline	PyAnnote + OpenAI
Diarization Accuracy	Good	Excellent
Setup Complexity	Simple	Moderate
Processing Speed	Fast	Slower
Cost	Lower	Higher
Best For	Quick processing	High accuracy needs

🧠 Technical Details

OpenAI Pipeline

• Uses OpenAI Whisper for end-to-end processing
• Includes audio preprocessing (noise reduction, VAD)
• Supports multiple audio formats
• Generates comprehensive meeting summaries

PyAnnote Pipeline

• Advanced speaker diarization using PyAnnote.audio
• Voice Activity Detection (VAD) for precise segmentation
• OpenAI Whisper for accurate transcription
• Word-level timestamp precision

🤝 Contributing

1. Fork the repository
2. Create your feature branch (git checkout -b feature/AmazingFeature)
3. Commit your changes (git commit -m 'Add some AmazingFeature')
4. Push to the branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgments

• OpenAI Whisper for transcription capabilities
• PyAnnote.audio for speaker diarization
• Hugging Face for model hosting and APIs

📞 Support

If you encounter any issues or have questions, please open an issue on GitHub.

---