Automatic profiling
Detects variable types, normality and dependencies from a reference CSV.
Synthetic data SaaS platform
Create secure synthetic data, validate it and share it without exposing sensitive information
SynthGAN turns statistical models into interoperable data products to train AI, simulate scenarios and participate in data spaces with privacy, traceability and governance.
- Privacy by design
- Built-in statistical validation
- Interoperable metadata and exports
From real data to synthetic product
Multivariate latent model
Simulates correlations, partial correlations or betas and transforms to continuous, binary, ordinal and categorical variables.
Validation and privacy
Compares real vs synthetic distributions, measures rarity and exposes non-specificity metrics.
Traceable export
Generates CSV, Parquet and JSON metadata ready for catalogs, APIs and data spaces.
Engine capabilities
Supported variable types
Dependency parameterizations
Reproducible with fixed seed
End-to-end pipeline
Use cases
Vertical segmentation according to the project communication plan.
Health and longevity
Synthetic cohorts for research without exposing real clinical data.
Public administration
Policy simulation and planning with secure aggregated data.
Data spaces
Interoperable products with metadata for catalogs and hubs.
AI teams
Training, QA and validation with non-sensitive datasets.
Institutional backing
SynthGAN is supported by European and Spanish public institutions.
SynthGAN is a project funded by the Spanish Ministry for Digital Transformation and Civil Service through project TSI-100130-2024-0151, financed by European funds from the Recovery, Transformation and Resilience Plan and "Funded by the European Union – NextGenerationEU"