RAG
Retrieval-Augmented Generation — grounding language model output in retrieved evidence rather than parametric memory.
Retrieval-Augmented Generation (RAG) is an architectural pattern where a language model's output is grounded in documents or data retrieved at query time, rather than relying on what the model memorized during training. In practice, RAG is one of the most useful ways to make Artificial Intelligence systems reliable enough for high-consequence work. The value of RAG is not fluency — it's traceability. A well-designed RAG system can cite its sources, abstain when evidence is thin, and tell you what it retrieved and why. That matters most in domains where being confidently wrong is worse than being usefully uncertain.