Skip to content

Python RAG Framework

Python RAG Framework is a lightweight library for building local retrieval-augmented generation workflows with clean, typed Python APIs.

The project focuses on a small set of capabilities:

  • document loading
  • deterministic chunking
  • local embeddings
  • semantic retrieval
  • prompt construction
  • pluggable LLM integrations

It is intentionally not a LangChain clone. The goal is a maintainable portfolio library with clear architecture and predictable behavior.