Glossary · Fundamentals
Retrieval-Augmented Generation (RAG)
A technique where an AI model retrieves relevant documents at query time and uses them to generate a more grounded, up-to-date answer.
Why it exists
A model's training data has a cutoff and can't include everything. RAG lets an engine pull in current, specific information — like a recent article or product page — at the moment someone asks a question, rather than relying purely on what it learned during training.
Relevance to AI visibility
Engines that use RAG (Perplexity is a clear example) tend to cite the sources they retrieved directly in the answer, which makes the underlying source material itself a visible, trackable part of the response.
See more terms in the full glossary, or read the AI Visibility 101 guide for the full picture.