Transforms and Synthetics on Relational Databases

The text discusses two new Gretel notebooks: multi-table transform and multi-table synthetics. These can be used independently or together to anonymize data in a relational database while maintaining referential integrity of primary and foreign keys. The multi-table transform removes personally identifiable information (PII) from the database, while the multi-table synthetics adds another layer of protection by generating synthetic data that maintains statistical properties of the original dataset. Both notebooks are demonstrated using a mock ecommerce relational database. Users can follow along with Gretel's multi-table notebooks available on their GitHub repository.