DataCebo Released SDV 2.0 for Synthetic Data Generation

The platform enables enterprise synthetic data creation within private environments to enhance testing and development workflows.

Updated on Oct. 2, 2026 in Artificial Intelligence

Bold vector editorial illustration of interlocking colored cubes connected by rods, symbolizing complex data relationships and synthetic structure. — DataCebo launched SDV 2.0, a synthetic data generation platform designed to allow enterprises to create non-sensitive test data within private environments. AI Illustration. Upload story photo >

Live Poll

Would you use synthetic data to test internal software instead of your actual company data?

DataCebo has launched SDV 2.0, a software tool designed to generate synthetic relational data directly within a customer's computing environment. The tool, which represents a commercial evolution of a framework with over 18 million downloads, automates the complex processes of schema detection and constraint enforcement.

Why it matters

This release addresses the operational burden and security risks associated with moving production data for development and testing. By enabling localized, synthetic data generation, companies in regulated sectors can maintain data privacy while ensuring their testing environments remain representative.

The SDV 2.0 software operates on standard central processing units, bypassing the need for GPU clusters. It supports direct integration with Oracle, SQL Server, BigQuery, Spanner, and AlloyDB at a starting price of $500 per month.

The players

DataCebo

A Cambridge-based company that develops tools for synthetic data generation derived from MIT research.

ING Belgium

A financial services institution that utilized the SDV framework to generate 10,000 synthetic payments.

Epiconcept

A research and development firm that utilized SDV software to construct a synthetic database in 55 minutes.

The details

The SDV 2.0 software functions by learning the underlying data relationships, structures, and business constraints from a representative subset of a company's actual production database. It uses these learned parameters to synthesize new, non-sensitive records that mimic the original dataset's statistical properties. The tool automates schema detection and constraint enforcement, effectively mapping how different tables and columns link together to maintain logical integrity in the synthetic output.

DataCebo launched the SDV 2.0 software on October 2, 2026.

The Tech Race

The release of SDV 2.0 marks a commercial transition from the original open-source SDV framework, which has been cited in 5,000 research papers. This move signals a competitive shift toward enterprise-grade, localized synthetic data tools that prioritize deployment within internal environments.

Enterprise data teams can immediately deploy the software within their own compute infrastructure to replace manual data masking tasks. Users should note that the $500 monthly subscription model requires existing database connections to systems like BigQuery or SQL Server.

The takeaway

The transition to SDV 2.0 highlights the industry shift toward privacy-first synthetic data pipelines for high-stakes development. Watch for future performance benchmarks comparing synthetic database training times on standard CPUs against previous-generation masking tools.