Retrieval-augmented generation (RAG)
RAG is a method in which a language model retrieves relevant documents from your own knowledge base before answering.
A language model on its own only knows what was in its training data — not your contracts, policies or product data. RAG closes this gap: for each question, matching passages from your documents are retrieved first and passed to the model as context.
The practical benefit is twofold. First, answers become verifiable because the sources used can be cited. Second, the risk of invented information drops considerably, because the model answers from concrete material rather than from memory.
This requires a clean knowledge base. Outdated or contradictory documents lead to wrong answers even with technically flawless RAG — the quality of the sources determines the quality of the results.
Related service
Related terms
Let's talk about your project.
Free initial consultation, 30–45 minutes, remote. An honest assessment — even if the answer is that you don't actually need it.