Retrieval-augmented generation (RAG)
Definition
Retrieval-augmented generation is a design in which an AI model first retrieves relevant passages from trusted documents or databases and then answers using them, rather than relying only on what it learned in training.
Why It Matters
It is the most common way enterprises ground AI answers in their own policies and data, and it makes answers easier to check.
More in Enterprise AI and Agents.
Related Terms
Laurenceo Intelligence is the daily AI intelligence briefing published in Bermuda by Laurenceo, founded by Jorai Laurenceo. Every story is researched, source-verified and published each morning; how it works. General information only, not legal, regulatory or financial advice.