Synthetic data disclosure

All data on this site is synthetic. No PHI. No production claims data.

What "synthetic" means here

Every claim, patient, provider, device, reading, and dollar amount shown anywhere on this site comes from a generated test corpus of 521,997 synthetic claims across 2,000 synthetic providers. The corpus is produced by a seeded, deterministic generator and is byte-reproducible. No record in it corresponds to a real person or a real billing event. "Mrs. Garcia" is not a real patient.

About the precision figures

Every precision number published on this site is a measurement against labeled synthetic ground truth: fraud archetypes injected into the synthetic corpus with known labels. These figures describe how the rules perform on that corpus. They are not field performance, and we do not present them as such. Performance on real data is established through a calibration pilot on the customer's own claims, and reported to that customer.

No PHI

The system contains no protected health information. Demo bundles are built from synthetic records, and public timestamp anchors only ever receive a cryptographic hash, never record contents.

Forward-looking statements

This site describes a product at prototype stage and plans that may change: calibration pilots, phased go-to-market, and pending patent filings. Statements about future capability, availability, or performance are intentions, not promises. Where a legal posture is described, such as suitability of evidence for expert scrutiny, it is a design goal, not a legal conclusion or a prediction about any specific proceeding.