rag-zero-to-hero-guide

Why we included this project

Anyone moving from a vague idea of retrieval-augmented generation to actually building RAG systems will find a structured, self-paced path here. The guide explains what RAG is, why LLMs need it, and how indexing, retrieval, augmentation, and generation fit together, then offers runnable Jupyter notebooks that build the same pipeline from scratch, with LangChain, over websites and YouTube transcripts, and as an agentic system on CrewAI. The RAG evaluation course is the standout part: it shows how to measure retriever and generator quality with RAGAS, DeepEval, and hallucination detection, which is exactly what people overlook when starting out. Developers and tech leads can use the toolkit section as a curated map of frameworks, document parsers, vector databases, and chunking libraries to shortcut their own tool selection. Treat this as a learning resource and reference rather than deployable software, the book you keep open while prototyping.