Glossary

RAG (Retrieval-Augmented Generation)

RAG is a method where the AI first looks up the relevant parts of your documents, then writes its answer using only those parts.

RAG stands for Retrieval-Augmented Generation. That is a heavy name for a simple idea. First, retrieval: when a question comes in, the system searches your documents and pulls out the few parts that match. Second, generation: the AI writes a reply using those parts. It is like an open-book exam, where the student may only answer from the pages in front of them.

Why do it this way? Because a general AI knows a lot about the world but nothing about your shop. RAG gives it your pages at the moment it needs them. It also keeps the AI honest: if the search finds nothing relevant, there is nothing to answer from, so the right reply is 'I do not know'.

This is the idea behind every fad. When a visitor asks something, your fad looks through the sources you fed it, picks the matching parts, and answers from those parts only. You do not have to set anything up. Feed good sources and the open-book exam goes well.