Appearance

Vector Search ​

Vector Search enables searching data using natural language queries. We can enable the Vector Field for your fields properties, which will automatically calculate embeddings and generate vector query params. Currently, Vector Search can only be enabled for #TypeString and cannot be used in conjunction with nested or array types.

typescript
import { db } from "@tailor-platform/sdk";

export const product = db.type("Product", "Product model", {
  // ...
  description: db.string().description("Product description").vector(),
  ...db.fields.timestamps(),
});
export type product = typeof product;

We can use Vector Search specifying vector: {<field name>: "natural language query"}, which returns results in the order they match the query:

graphql
query VectorSearchProduct {
  products(vector: { description: "What products are associated with A and B?" })
}

It is useful to specify the first and to use a query filter to narrow down the results with Vector Search.

graphql
query VectorSearchProduct {
  products(
    first: 10
    query: { name: { contains: "XX" } }
    vector: { description: "What products are associated with A and B?" }
  )
}

Function Service Support ​

Vector Search is also available in the Function service using SQL syntax. You can use the vector_similarity function to sort records by similarity. For more details, see Accessing TailorDB from Function service.