CryptoRAG

RAG-powered AI chatbot for cryptography education.

This project was inspired by the course CPS 633.

Note: Recently migrated from a Streamlit app to a React + FastAPI app.

Uses:

• Frontend: Next.js, Tailwind CSS, Framer Motion
• Backend: FastAPI, Python
• AI/RAG: Cohere (Command R+, Embed v3, Rerank v3), Unstructured, PDFMiner

Usage

Backend

Maneuver to the api directory and start the server:

uvicorn api.index:app --reload

Frontend

Maneuver to the frontend directory and start the client:

npm run dev

The app will ingest sources (including Crypto 101) on startup and be ready for queries.