What is a Vector Database?

A vector database is a type of database that provides a way to store representations of information — ”vectors” — that make it fast to compare relationships between each. Most modern AI search relies in part on vector search and implements some form of vector storage & retrieval.

Each vector in a vector database represents an object. It can be something as simple as a string of text, or as complex as an image, or any other datatype that can be turned into bits. Generally, each vector is created by applying an ‘embedding’ function to the original object, which creates a series of dimensions — a vector can have a handful of dimensions or thousands. Each dimension defines that vector’s relationship to other vectors that share that dimension.

Do I need a vector database?

For specialized use cases, you may need to host your own vector database. And you’ll need to connect it with a bunch of other piece-parts to create the end-to-end search experience you’re imagining.

But Objective handles all of that for you - you just add Objects to your Object Store, and create the kind of Index (text, image, or multimodal) for the kinds of search you want to perform. Objective goes to work behind the scenes to create the representations of your data needed to give you modern, lightning-fast AI-native search.

We love AI search infrastructure, and we’re constantly finding & integrating the best of the best. We’ll handle the messy search infrastructure, you go build the next great search experience!