Retrieve
Relevant passages are found from a document collection using keywords, embeddings, or a hybrid search strategy.
Retrieval-Augmented Generation
Retrieval-Augmented Generation (RAG) combines search with language generation so an AI system can ground an answer in a selected collection of reference material.
Augment
The selected context is placed alongside the user question to give the model a focused evidence window.
Generate
A language model uses the question and retrieved context to produce a clear, useful response with better traceability.