Glossary · RAG

Retrieval-Augmented Generation (RAG)

Retrieval-augmented generation (RAG) connects a language model to an external knowledge source, retrieving relevant information at query time and supplying it as context — so answers are grounded in real, current data rather than only training.

RAG is the default pattern for enterprise AI that needs accurate, current, source-attributable answers. It updates by re-indexing (no retraining), keeps proprietary data in your own store, and supports citations.

It is not foolproof: retrieval quality, chunking, and evaluation determine whether the right passage is found and actually used.

Related terms

Agentic RAG
Embedding
Vector database
Large Language Model (LLM)
Hallucination

Learn more

What is RAG?
RAG vs fine-tuning

Sources

Lewis et al. (2020), Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks (arXiv:2005.11401)

Book a demo Back to glossary