Simple query

Now that we have our database set up, tables tracked and remote relationship defined, we are ready to start querying. This is where we start to see power of the Hasura integration.

Let's say I wanted to access Resumes and search over the vectorized near_text property. I can do this easily with Hasura's GraphQL API. Head to the API tab and run this query:

Copy
query SimpleQuery {
  Resume(where: { vector: { near_text: "marketing experience" } }) {
    application_id
    content
    vector
  }
}

You should see a list of results come through as a response on the right-hand side of the explorer. The near_text property is a vectorized field. Vector databases, such as Weaviate, store information as vectors in what are called embeddings. These embeddings are a mathematical representation of the data. This allows us to perform similarity searches over the data. In this case, we are searching for resumes that are similar to the phrase "marketing experience".

Our vector database is literally storing similar information near each other in the database!