Skip to main content
The Data Validation agent helps you set up validations without manually navigating configuration options. Describe what you want to check — the agent identifies the target dataset and validation type, then generates the appropriate validation configuration for your approval.

​
What it does

  • Identifies the target dataset from your natural language description
  • Determines the appropriate validation type
  • Generates validation configurations with a description and confidence score
  • Waits for your approval before making tests available to run

​
Create an agent

  1. Go to the Agents tab in the left sidebar.
  2. Click Create Agent.
  3. Select Validation Generation as the agent type.
  4. Select the data source, and add a name and description for the agent.
  5. Optionally, add additional instructions to guide the agent’s behavior (for example, focus areas or validation preferences).
  6. Click Create to save.

​
Run an agent

  1. Go to the Agents tab and click the agent you created.
  2. Select the table you want the agent to run on.
  3. Click Run. You can create multiple runs for the same agent, each targeting a different table.
  4. Once the run completes, the agent generates a list of validations — each with a description and a confidence score.
  5. Review each validation and approve the ones you want. Only approved validations become tests that are ready to run.

​
Supported validation types

The agent can configure the following validation types:
TypeDescription
NOT NULLChecks that a column contains no null values
UNIQUEChecks that all values in a column are distinct
CHECKChecks that column values satisfy a condition
FOREIGN KEYChecks referential integrity between tables
Custom SQLValidates data using a custom SQL assertion
Delta ValidationChecks the change between two time periods
For a full list of supported validation types and their configuration options, see Validation Types.