Generators define how synthetic values are produced for columns in Synthetic Data. A generator encapsulates the logic used to generate values such as dates, sequences, or formatted strings so the same generation behavior can be reused consistently across datasets and jobs.
Each generator is based on a framework and can include classification rules that control how it is automatically matched to columns during discovery. These rules use table and column name patterns to reduce manual assignment and ensure consistent data generation.
Generators are managed independently of datasets and jobs. Changes to a generator take effect the next time a job that uses it runs.
Next steps
Last updated
Oct 8, 2026