CitenzoRetrieval-Augmented Generation (RAG)
A technique where an AI model retrieves relevant documents at answer time and generates its response from them — the basis for cited AI answers.
Retrieval-augmented generation (RAG) is a method where, instead of answering purely from memory, an AI system first retrieves relevant documents and then generates its answer from them. It's the architecture behind search-connected assistants that can cite sources.
For GEO, RAG is why third-party presence matters so much: the model's answer is only as good as what it retrieves, so being in the documents it pulls is what gets you mentioned.