A census of people who never existed. 65 attributes per record — demographics, health, financial, behavioral — drawn from published US reference distributions, calibrated to NHANES & ACS. Not real data. Statistically indistinguishable from it.
curl -s -X POST https://api.simpleidgen.com/v1/datasets/person \
-H 'Content-Type: application/json' \
-d '{"clientId":"…","count":100000}'Marginals are drawn from published US references — ACS, NHANES, CDC NDSS, KFF, MEPS, BLS. BMI follows weight over height squared. ZIP codes match states. Diabetes prevalence rises with age. Verified with KS and chi-squared tests.
Most synthetic-data tools need your real records first, then learn from them. This one doesn't. It is built entirely from public reference data — no real PII ever enters the system at any point.
Same seed, same people — byte for byte, every time. Pin your test fixtures, reproduce a bug, version a dataset. Change the seed and you get a different, equally-calibrated population.
Need millions of rows? Submit a job and the async endpoint streams up to 5,000,000 records as JSONL to object storage. Poll for status, then download. The link stays live for 90 days.
Demo to investors with records that read like real patients, not John Doe, 30. Build your MVP before you have a single user.
A 10K-row dataset in five seconds instead of six months fighting an IRB. Teach analytics, prototype models, publish reproducibly.
Populate staging with believable data. Deterministic seeds make every fixture reproducible across CI runs and machines.
A full example record and all 65 fields on the product page. Create a free account to generate your own.