Knowledge-base search (RAG)

Retrieval-augmented generation (RAG) is the technique behind assistants that answer from your data: the system retrieves the relevant records — artworks, contacts, notes, correspondence — and the model composes an answer grounded in what it found, rather than inventing from general training.

For a gallery, the payoff is one search across everything: "what did we quote Berger for the woodcut?" resolves against mail, offers, and records together — with the sources shown.