What is RAG?
RAG is the technique of fetching relevant documents from your own data and giving them to a language model alongside the question, so its answer is grounded in your facts.
A model on its own knows nothing about your price list, your policies or last week's orders. RAG looks those up first, from a search index over your documents, and puts them in front of the model. The answer then cites what it was given.
It is how a support assistant answers from your actual documentation rather than inventing. Its quality depends almost entirely on the retrieval: what gets indexed, how it is chunked, and whether the right passage is found.
A RAG system is only as current as its index, so the ingestion of new documents has to be automated or the assistant slowly starts answering from last year.