RAG Architecture

RAG architecture describes the components and data flow used to retrieve external information and provide it to a generative model before it produces an answer. The concept matters when teams build production generative-AI applications that need reliable context, controlled outputs, evaluation, and connections to business data.