AI & Vectors

Collections


A collection is an group of vector records. Records can be added to or updated in a collection. Collections can be queried at any time, but should be indexed for scalable query performance.

Each vector record has the form:

1
Record (
2
id: String
3
vec: Numeric[]
4
metadata: JSON
5
)

For example:

1
("vec1", [0.1, 0.2, 0.3], {"year": 1990})

Underneath every vecs collection is a Postgres table

1
create table <collection_name> (
2
id string primary key,
3
vec vector(<dimension>),
4
metadata jsonb
5
)

where rows in the table map 1:1 with vecs vector records.

It is safe to select collection tables from outside the vecs client but issuing DDL is not recommended.