A vector database stores text as numbers, so AI can find documents with the same meaning. Domino IQ has a built in vector database from Notes 2026.
A vector database stores text as long rows of numbers. The numbers are called vectors. They describe what the text is about. The work of turning text into numbers is called embedding.
Why numbers?
An ordinary full text index finds the same words. A vector database finds the same meaning. If you ask about "holiday", it can also find a document about "time off in July". That makes it possible to ask a whole question in plain words.
The vector database is the basis for RAG. It finds the right documents before the language model answers.
The vector database in Domino IQ
From Notes 2026 a vector database comes with Domino IQ. It needs no new ODS. It sits as a file in the Domino data folder next to the Notes database, a bit like a full text index. Each database has its own file.
The file holds numbers, not readable text. It stays on the local server. Domino IQ uses a special embedding model to make the numbers.
