The engine's output, on a portfolio you can browse.
A simulated 2,040-account portfolio with planted synthetic identities, reviewed by the actual DoubleCheck engine. Every disposition and reason code below is real engine output on this simulated book.
Planted fraud separates cleanly here, so this shows the mechanism, not real-world performance. No institution's data appears on this page, and none ever will.
Sample of accounts. The interactive explorer needs JavaScript.
| Account | Disposition | Opened | Type | Fired signals |
|---|---|---|---|---|
| sa_00000 | Fail | 2022-03-11 | checking | provenance, coherence, inflection, ring |
| sa_00001 | Fail | 2022-03-16 | checking | provenance, coherence, ring |
| sa_00002 | Fail | 2023-02-23 | checking | provenance, coherence |
| a_0000000 | Pass | 2023-07-09 | checking | none |
| a_0000001 | Pass | 2023-05-26 | checking | none |
| a_0000002 | Pass | 2022-11-16 | checking | none |
Dispositions are pass, review, or fail with reason codes. No scores and no dollar figures. Click any account, including a pass, to see every value the engine measured on it and the threshold each was compared against. The series shown are utilization percent, monthly counts and shares, never amounts.