RAG, or Retrieval-Augmented Generation, is the pattern of retrieving relevant source material before an LLM generates an answer.
RAG improves groundedness by pulling in current documentation, articles, code, or internal knowledge before the model answers. This is especially useful when the information changes faster than the model can be retrained.
It is one of the most common ways to reduce hallucinations in product and support workflows.
RAG does not automatically remember users across sessions. It can retrieve documents, but it is not the same thing as a durable memory layer for preferences, instructions, or prior decisions.
That is why many modern agent stacks pair RAG with memory rather than choosing one or the other.
Use these glossary pages and commercial landing pages to move from definition to implementation.