What is Retrieval-Augmented Generation?
Combining a search step with a language model so answers are grounded in fetched data.
Formal Definition
Retrieval-augmented generation (RAG) pairs an information-retrieval system with a generative model: relevant documents are fetched first, then supplied to the model so its output is grounded in current, specific data rather than memory alone. It improves accuracy and recency. The pattern mirrors how a market model reasons over freshly retrieved news and fundamentals.
In Simple Terms
It is giving an AI an open book before it answers. Rather than relying only on what it memorized during training, the system looks up relevant, up-to-date facts and hands them over so the answer is grounded.
Example
Feeding a model the latest headlines and price data for its holdings before it decides is a retrieval-augmented approach: the decision rests on fresh evidence, not stale memory.