Record

A record is one row in a vector table. It represents one retrievable unit and consists of a record ID, a dense vector, and metadata.

Record ID

A record ID is the unique identifier for the given record.

Record IDs can be provided manually or generated automatically if the vector table is created with auto_generate_id set to true.

Dense Vector

A vector, sometimes called an embedding, is an ordered list of numeric values that captures the meaning of a piece of data. You can think of each value as a coordinate in a high-dimensional space. Items with vectors that are close together in space tend to be similar in meaning. Only vectors with a dense storage format are supported with Autonomous AI Vector Database. With dense vectors, every dimension stores a value, including zero-values.

To produce dense vectors, you run your data through an embedding model. Depending on the workflow you choose, the embedding can be performed automatically by Autonomous AI Vector Database using an integrated embedding model, or you can bring your own, in which case you perform the embedding with an external embedding model and then upsert the resulting vectors.

Metadata

Metadata is structured information that is stored alongside vectors. It enables the following:

Autonomous AI Vector Database APIs support querying using vector similarity plus metadata filtering in the same retrieval flow.