Python
LangChain
Docker
FastAPI
Qdrant
RAG Starter Kit
Open-source boilerplate for quickly bootstrapping production-grade RAG systems.
Overview
A batteries-included starter kit for building RAG systems without the boilerplate. Includes ingestion, retrieval, evaluation, and a simple chat UI out of the box.
What's included
- Ingestion pipeline with chunking strategies and metadata extraction
- Vector store integration (Qdrant by default, swappable)
- FastAPI backend with OpenAI-compatible
/chatendpoint - Evaluation harness using RAGAs metrics
- Docker Compose for local development
- GitHub Actions CI pipeline
Getting started
Clone the repo and run docker compose up. The ingestion pipeline, vector store, and API are ready in under a minute.