A Field Guide to RAG: 42 Runnable Techniques

What it does

This repository collects Jupyter notebooks that implement specific Retrieval-Augmented Generation patterns, from foundational chunking to graph-based retrieval. Each notebook demonstrates a single technique—such as MemoRAG, JSON RAG, or Open-RAG-Eval—so you can inspect the mechanics without building a pipeline from scratch. The author pairs the code with external write-ups that explain the intuition behind each approach and when it quietly fails.

The interesting bit

Instead of treating RAG as a solved problem with one true architecture, the repo frames it as a design space of 42 distinct strategies you can compare side-by-side. That makes it less of a framework and more of a field guide for retrieval mechanics.

Key highlights

• 42 runnable notebooks covering foundational to advanced RAG patterns
• Recently added techniques include MemoRAG, Open-RAG-Eval, end-to-end evaluation, and JSON RAG
• Each notebook is accompanied by a detailed explainer on when the technique wins (and when it quietly fails)
• Links out to related repos covering agents, prompt engineering, and agent memory
• Actively accepts community contributions with documented guidelines

Verdict

Developers who are debugging retrieval pipelines or choosing between chunking and reranking strategies will find this a useful survey. Skip it if you need a single, opinionated production framework rather than a broad technique catalog.