This project is a powerful tool that harnesses the of OpenAI's Whisper ASR and GPT-4 models to automate the process of generating meeting minutes from audio recordings, as well as text summary and key point extraction.

This Python-based application transcribes audio, provides summaries, extracts key points and action items, and performs sentiment analysis.

You should have a basic understanding of Python and an OpenAI API key. You can get an API key here. This project was built using Python 3.11.5.

Create a virtual environment and install the requirements:

python -m venv venv
source venv/bin/activate
pip install openai
pip install python-docx
pip install python-dotenv

Add your OpenAI API key to the .env file.

1. Transcribe Audio with Whisper ASR. Refer to transcription.py
2. Summarize and Analyze Transcript with GPT-3. Refer to sentiment_analysis.py
3. Extract Key Points and Action Items. Refer to sentiment_analysis.py
4. Generate Meeting Minutes. Refer to meeting_minutes.py
5. Save Meeting Minutes as a Word Document. Refer to save_as_docx.py

• Improve the accuracy of the Whisper ASR model by training it on a custom dataset.
• Experiment with prompt engineering to enhance model performance.
• Customize the system messages to provide more context to the GPT-4 model for better results.

MIT

• OpenAI API