Test Data
Overview
Test Data Generation lets you create schema-aware datasets for testing without copying production data or maintaining manual fixtures. Generated data conforms to the detected schema (types, formats, constraints), so you can run meaningful tests without exposing sensitive information. This guide explains how to generate test data for a single table or across related tables while keeping relationships consistent, and how to download and reuse generated datasets in local development, staging, and CI pipelines.
The test data generation process consists of the following steps:

Test data tab is only visible when Test Data Feature is enabled for the platform and the selected datasets metadata includes semantic types.

The form lets you provide details to identify your test data, and also configure generation parameters. The form has the following fields:
The name of the test data.
A description of the test data for all other users to see.
A read-only dropdown input specifying the table containing the subjects for the test data generation. This pre-filled data is automatically extracted from the dataset's metadata and cannot be edited during test data generation.
The number of subjects to be generated in test data.
The format in which you want the test data to be generated.

To preview metadata used during a test data generation, click on Metadata Snapshot button on the right. It should open a pop-up window which displays the metadata snapshot associated with this test data.
