RAG Project

A small Retrieval-Augmented Generation (RAG) project that:

• loads PDF files
• chunks PDF text
• generates embeddings with sentence-transformers
• stores vectors in chromadb
• retrieves the most relevant chunks for a query
• uses retrieved context for a simple RAG prompt

Project structure

• pyproject.toml — project dependencies and Python settings
• requirements.txt — pinned dependency list
• notebook/document.ipynb — main pipeline notebook
• data/pdf/ — source PDF files
• data/vector_store/ — persistent Chroma vector store

Prerequisites

• macOS Intel (x86_64)
• Python 3.11.x (recommended)
• uv package manager installed
• A valid Groq API key if you want LLM answers via Groq

Install

cd RAG
uv add -r requirements.txt