Learn RAG
New to Retrieval-Augmented Generation? Start here. These guides explain the core ideas in plain English, no prior background assumed. When you hit a term you do not know, the glossaryhas a friendly definition.
Retrieval-Augmented Generation lets an AI system answer from documents you trust rather than from memory alone. These guides build the idea up one piece at a time, so you can start from zero and finish ready to build something real.
A sensible path: begin with What is RAG?for the whole picture, and What is a dataset?if the raw material is new to you. Then take the three ideas the pipeline rests on, explained in plain English: embeddings,chunking, andvector databases. Before you build on real data, data licensingcovers what you are, and are not, allowed to do with it.
Prefer to learn by doing? The hands-on Build your own RAG from scratchtutorial walks through every stage in Python in about an hour. And whenever a term is unfamiliar, the glossaryhas a short, friendly definition. Everything here is free and assumes no background; when you are ready to find data, the directory is waiting.
Courses
Guides
What is RAG?
How Retrieval-Augmented Generation works, and why it beats asking a model to remember everything.
Embeddings explained
What embeddings are, how they turn text into meaning you can search, and how to choose an embedding model.
Chunking your data
Why you split documents into chunks before retrieval, and how to do it well.
Vector databases 101
What a vector database does, how nearest-neighbour search works, and how to pick one.
Data licensing for RAG
A practical guide to licences you will meet, and what they let you actually do.
What is a dataset?
What a dataset actually is, the forms it takes, and what separates a useful one from a liability.
Glossary
Every bit of jargon you will meet in RAG, explained simply. 42terms and counting.