Tired of manually taking notes during meetings? This project automates the process of identifying actionable items from a transcript, saving you time and ensuring no action is missed.
The pipeline processes the transcript through a series of steps to ensure accurate and reliable extraction.
- Chunking: Raw transcripts are split into smaller, manageable pieces to fit within the language model's context window.
- Clustering: Similar content is grouped together to handle repeated discussions or actions.
- Action Filtering: Non-actionable segments are filtered out early to improve efficiency.
- Extraction (Chain of Thought): Action items are precisely extracted using a reliable, prompt-based reasoning approach.
This section will guide you through the process of setting up and running the project using Docker.
-
Clone the repository.
-
Build the Docker image:
sudo docker build -t your-image-name path-to-dockerfile
-
Run the container with your API key:
sudo docker run -p 8000:8000 --name your-container-name -e GEMINI_API_KEY="your_api_key_here" your-image-nameReplace
"your_api_key_here"with your actual API key.
Once the container is running, the API will be accessible on your local machine.
Open your browser and navigate to http://localhost:8000/docs to access the interactive API documentation provided by FastAPI.
You can test the API by sending a POST request with a transcript in the request body. A sample transcript is available in examples/transcript.json.