Usage

The Semantic Search service can be used through a REST API. You would need to consume it by sending HTTP requests to the different service endpoints. There is an endpoint for every action available in the service.

OpenAPI documentation

The REST API has been developed following the OpenAPI specification. This means that the REST API complies with the OpenAPI standard and that you can use the OpenAPI documentation to learn how to use the API.

In fact, the Semantic Search service provides an endpoint where you can find the OpenAPI documentation. You can find it at http://localhost:8000/docs.

This endpoint will open a Swagger UI page where you can find all the information about the API. You can also try the API from this page, as it provides a way to send requests to the different endpoints.

Additionally, you can get the openapi.json file from the /docs endpoint. You can find it at http://localhost:8000/openapi.json.

Available endpoints

The Semantic Search service provides two endpoints: learn and search. You can find more information about them in the following sections.

Learn endpoint

With the learn endpoint, you can make your service learn from a set of documents' content (or any text you'd like!). You just need to provide the endpoint with content, so it can then be turned into embeddings and stored in the local database.

You can try the service by sending a POST request to http://localhost:8000/learn.

JSON Body

{
    "items": [
      {
        "content": { "text": "Grain in dogs food is not good for them."},      
        "metadata": {"key1": "value1", "key2": "value2"},
        "cluster_id": "your_file_id"
      }
    ]
}

CURL Request

  curl --location 'http://127.0.0.1:8000/learn' \
  --header 'Content-Type: application/json' \
  --data '{
    "items": [
      {
        "content": { "text": "Grain in dogs food is not good for them."},      
        "metadata": {"key1": "value1", "key2": "value2"},
        "cluster_id": "your_file_id"
      }
    ]
  }'

Search endpoint

With the search endpoint, you will get answers to your questions, based on what you have ingested with the learn endpoint. You just need to provide the endpoint with a query, and the service will return the most relevant information based on the embeddings stored in the local database.

You can try this endpoint by sending a POST request to http://localhost:8000/search with the following body:

JSON Body

{
  "query": "How does grain food affect dogs?",
  "cluster_ids": ["your_file_id"]
}

CURL Request

  curl --location 'http://127.0.0.1:8000/search' \
  --header 'Content-Type: application/json' \
  --data '{
  "query": "How does grain food affect dogs?",
  "cluster_ids": ["your_file_id"]
  }'