A Retrieval-Augmented Generation (RAG) system that automatically processes text documents from Google Drive, stores embeddings in PostgreSQL, and provides a question-answering interface.
- Automatic document processing from Google Drive
- Vector similarity search using pgvector
- Real-time question answering through Gradio interface
- Daily automated checks for new documents via GitHub Actions
- Supabase account (for PostgreSQL database)
- Google Cloud account with Drive API enabled
- Python 3.10 or higher
- Develop Gradio app for RAG-enabled chat
- Deploy chat app using Docker and Google Cloud
- Continue adding data