easy-local-rag

Why we included this project

Easy-local-rag runs retrieval-augmented generation entirely on your own machine, using Ollama to run local models and mxbai-embed-large to handle embeddings. You upload PDF, text, or JSON files and ask questions, and the same flow can be pointed at your own email inbox when you want to search your correspondence. The project is a set of plain Python scripts, so every stage from document ingestion to query rewriting stays readable, which makes it a good way to learn how RAG fits together instead of treating it as a black box. The companion YouTube tutorials walk through the setup step by step, which helps developers and small teams new to RAG get oriented quickly. It is not a production-grade retrieval system, but as a local proof of concept or a baseline to study, it gets you asking questions of your own data within minutes.