This documentation describes the integration of MindsDB with Vertex AI, a machine learning platform that lets you train and deploy ML models and AI applications, and customize large language models (LLMs) for use in AI-powered applications. The integration allows for the deployment of Vertex AI models within MindsDB, providing the models with access to data from various data sources.


Before proceeding, ensure the following prerequisites are met:

  1. Install MindsDB locally via Docker or Docker Desktop.
  2. To use Vertex AI within MindsDB, install the required dependencies following this instruction.


Create an AI engine from the Vertex AI handler.

This command creates a config object that can be used in client creation step.

CREATE ML_ENGINE vertex_engine
FROM vertex
      project_id = "mindsdb-401709",
      location = "us-central1",
      staging_bucket = "gs://my_staging_bucket",
      experiment = "my-experiment",
      experiment_description = "my experiment description",
      service_account = {
        <paste service account keys here>

Create a model using vertex_engine as an engine.

This command authenticates client to a Vertex account using config from previous step. If the endpoint for the model already exists, we create this model in MindsDB. Otherwise, we create and deploy the model to the endpoint before creating this model in MindsDB.

CREATE MODEL vertex_model
PREDICT target_column
      engine = 'vertex_engine',  -- engine name as created via CREATE ML_ENGINE
      model_name = 'model_name', -- choose one of models from your project
      custom_model = value;      -- indicate whether it is a custom model (True) or not (False)


The following usage examples utilize vertex_engine to create a model with the CREATE MODEL statement.

Detect anomaly using a custom model stored in Vertex AI.

CREATE MODEL vertex_model
    engine = 'vertex',
    model_name = 'diamonds_anomaly_detection',
    custom_model = True;

Query the model to get predictions by joining it with the data table.

SELECT d.cut, m.cut AS anomaly
FROM data_table as d
JOIN vertex_model as m;

Next Steps

Go to the Use Cases section to see more examples.>