Synthetic Data Generator by MOSTLY AI on Databricks Marketplace

This video demonstrates how to use the MOSTLY AI Solution Accelerator within Databricks to generate granular private synthetic data, which can help accelerate data access across an organization.

JB starts by guiding through accessing the MOSTLY AI Solution Accelerator via the Databricks Marketplace.

JB explains notebooks that make up the Solution Accelerator and outlines the process, which involves four key steps:

1. Training a Synthetic Data Generator: The first notebook allows to create a "Generator" that is used to create synthetic data that mimics real data without revealing sensitive information.

2. Saving Information to Unity Catalog: The second step involves saving important information and creating a discoverable model within Unity Catalog.

3. Generating Synthetic Data: The third step uses the model created in Unity Catalog to generate synthetic data.

4. Accessing the Generated Data: JB demonstrates accessing the generated synthetic data stored in a specified location within Databricks.