Upload a driver’s ELD report and supporting documents. Verity builds one standard timeline, checks it against 49 CFR Part 395, and queues every finding — evidence attached — for your auditor’s ruling.
ELD report plus fuel receipts, BOLs, weigh tickets, scanned or handwritten grid logs.
18+ ELD formats — Motive, Samsara, Verizon, and more — converted to one standard duty-status timeline.
A deterministic rules engine runs ~20 categories of HOS checks against the timeline.
The cockpit shows each possible violation with its times, events, and the evidence behind it.
A human auditor confirms or dismisses. The final compliance decision is never automated.
Compliance software shouldn’t be a black box. Verity’s engine is built to be repeatable, explainable, and held to the standard of named human auditors.
Fixed rules decide whether a violation occurred — never a model. Same logs in, same findings out, every time.
~90 hand-calibrated constants, each pinned to a specific real audit and a named auditor’s ruling, plus codified doctrine rulings from working compliance auditors. Measured at 91.1% precision and 90.4% recall against confirmed auditor answer keys from real staff audits — both over the 90% launch bar.
1,700+ automated tests run on every change. If a change would ever disagree with a human auditor’s confirmed key, it’s blocked automatically — fixing one audit can’t quietly break another.
Vision models read what parsers can’t — a photographed paper grid, a crumpled fuel receipt. Anything AI-read is labeled, scored for confidence, and marked “needs review” when weak. AI never decides compliance.
Every customer audit is graded against strict DOT settings — the standard an enforcement officer would hold you to — not a lenient internal mode.
No integration project required — export from your ELD portal and upload. Paper grid logs, scanned or handwritten, work too. Missing your vendor? Tell us — parsers ship fast.