LangChain RAG Starter Kit — ChromaDB + FastAPI + Streaming + Citations (Production-Ready)

Building a "chat with your documents" feature always starts the same way: you spend a weekend wiring together a document loader, a chunking strategy, a vector store, a retrieval chain, and a streaming API — before you've even gotten to the part that actually matters for your product.This kit skips that weekend.It's a real, working RAG backend — not a tutorial snippet — built on the same architecture used in a production AI compliance tool. Clone it, drop in your Anthropic API key, and you have a fully working doc-chat app running locally.What's included: 📄 Upload PDF / TXT / DOCX — automatically chunked and embedded ⚡ Token-by-token streaming responses (not a 5-second wait for the full answer) 🔗 Source citations on every answer — shows exactly which document it came from 💬 Multi-turn conversation memory 🗄️ Local vector store (ChromaDB) — no cloud database, no monthly bill 🖥️ A working chat UI included, so you can test it the moment you clone it 📘 Full README with setup steps, API reference, and a troubleshooting section ⚙️ Clean, swappable architecture — change the LLM, embeddings, or vector store without touching the rest of the pipeline Tech stack: FastAPI · LangChain · ChromaDB · Claude API · local embeddings (no OpenAI key required)Who this is for: Indie devs prototyping a "chat with your data" product Agencies building internal-docs or customer-support bots for clients Anyone tired of re-wiring the same RAG boilerplate for every new project Setup time: ~10 minutes, including installing dependencies and adding your API key. Tested from a clean environment — the README's steps actually work end to end.

Found on Devadex — the discovery index for independent software the big search engines bury. More from gumroad.

Report this listing