RAG over PDFs

This repo lets you index unstructured documents in a Chroma store, and then query them through an offline LLM served by Ollama. A simple UI is provided through streamlit.

Prerequisite

Install Ollama or run it with Docker
pull the needed models (e.g. ollama pull mistral)

Run the app

run poetry run python3 rag/docs_rag/index.py to index PDFs into a Chroma index.
run the chat with poetry run streamlit run rag/docs_rag/docs_chat.py