Definition. RAG (Retrieval-Augmented Generation) is a method in which a language model does not answer from training data alone. At runtime it fetches relevant information from an external knowledge source and includes it. Documents usually sit in a vector database. For each query, matching passages are retrieved and passed to the model as context.

Why it matters. RAG makes AI usable on a company's own, current content instead of the model's static training state. It reduces hallucinations and makes sources traceable. It is the basis for knowledge assistants and internal search. The prerequisite is structured, machine-usable content.

Related. MCP, Structured Content, AI-ready CMS, Agent Readiness