The Synthetic Data Playground
Pick a schema, drag the row count, generate a doppel in your browser. Nothing leaves this tab.
doppel_customers_25.csv
Quality report
signed · sha256:doppel-dStatistical fidelity
99.2%
Correlation match
0.985
Privacy budget (ε)
2.8
Memorisation risk
0.210%
What just happened?
The playground demo runs a tiny browser-side model — a few hundred lines of JavaScript. Production Doppelset runs an industrial-strength engine on your data. The principles are the same.
1
Schema is read
The picker hands the engine a JSON schema describing columns, types, and value ranges. In production we infer this from your DDL.
2
A doppel is trained
A seeded generator produces statistically-controlled distributions — gaussians, log-normals, categorical mixtures, joint relationships. In production: a hybrid tabular transformer.
3
Quality is reported
A signed report is generated alongside the data: fidelity, correlation, ε-spend, memorisation risk. Hand it to your auditor.
Ready to graduate from the playground?
Point Doppelset at your real schema. Generate billions of rows with audit-ready ε guarantees, multi-table consistency, and first-class connectors for the warehouse you already run.
Playground vs production
- browser modelindustrial engine
- 500 rows max4 B rows
- fixed schemasyour schema
- no ε proofsigned ε receipt
Ship faster. Stop arguing with legal.
Generate your first 100,000 synthetic rows in the next ten minutes. No credit card.