Glossary
A database purpose-built to store and search high-dimensional vector embeddings — numerical representations of meaning — enabling similarity search: finding content that's conceptually related rather than just keyword-matching.
Why It Matters
Traditional databases search for exact or partial text matches. Vector databases let a system find the most conceptually similar document, image, or passage to a given query, even if it doesn't share any of the same words — which is the retrieval mechanism behind most modern AI search and recommendation systems.
This is part of our AI Systems capability — the infrastructure that lets intelligence get embedded into a workflow, not bolted on as a standalone feature.
See the full AI Systems capabilityCan a regular database do this instead?
Some traditional databases have added vector search extensions, but purpose-built vector databases are generally optimized for the specific performance characteristics — speed and accuracy at scale — that similarity search over millions of embeddings requires.
What's actually stored in a vector database?
Numerical arrays (embeddings) that represent the meaning of a piece of content, usually alongside a reference back to the original source document or record.