RAG (Retrieval-Augmented Generation)

An architecture pattern where an AI agent retrieves relevant information from external knowledge sources before generating a response. Requires careful access control to ensure the agent only retrieves data it is authorized to see.