DataCebo Releases SDV 2.0 for Enterprise Synthetic Data

In a press release, DataCebo introduced SDV 2.0, a version of SDV Enterprise that builds generative relational models from enterprise databases within customer environments. The models learn statistical patterns, structures, relationships and business rules across tables without requiring production data to leave the organization.

SDV 2.0 can work from a representative data subset and model relational schemas of any depth. It automates schema and relationship detection, format identification, business constraint enforcement, model tuning and the creation of datasets for specific tasks. The software connects directly to Oracle, SQL Server, BigQuery, Spanner and AlloyDB.

Organizations can use the models to generate synthetic data, train and evaluate AI agents, and simulate scenarios and edge cases. SDV 2.0 is available now through self service plans, with consumption based pricing starting at $500 per month for unlimited tables.

The technology originated at MIT's Data to AI Lab. DataCebo cites customer results including ING Belgium generating 10,000 synthetic payments in two minutes, and says SDV has been downloaded more than 18 million times.