Every driver decision, reconstructable to the violation, the rule and the rulebook version.
Motor carriers and commercial auto underwriters answer for who they put behind the wheel. DQ Ledger turns the four records they already pull - CDLIS, the MVR, the PSP report and the Clearinghouse - into an approve, manual-review or adverse-action recommendation that shows its work, runs under a versioned rulebook, and routes every adverse outcome to a named person.
sample files land where a safety director expects, on both SMS severity scales - checked by a test that fails if a label and the engine disagree.
decisions change if Carrier hiring standard v3.3 replaces v3.2 - each with the rule responsible.
adverse result is a recommendation: it opens the FCRA pre-adverse and adverse-action steps and needs a named reviewer.
deliberately broken engines, each caught by the check written for it in 300 seeded trials.
Three layers, in order. Each can only raise the outcome, never lower it.
Bars and holds the regulations impose: a prohibited Clearinghouse status, a suspended CDL, an expired medical certificate, a missing endorsement, a required record not on file.
The choices a carrier or insurer makes above the floor - experience, major violations, crashes, out-of-service orders - versioned, with an effective date.
A risk index over the points ledger, using FMCSA SMS mechanics: severity, recency weights and a per-inspection cap. It catches volume no single rule sees.
No model touches a hire or insure decision. A model does the reading after an incident - and has to cite it.
- Verdict, binding rule and risk index
- The points ledger and its cross-record de-duplication
- Rulebook versions and their impact on the book
- The decision record and its input hash
Same input, same output, every time - which is what an audit, a bias test and a court need.
- Drafts a timeline, root causes and next steps from retrieved sources
- Every claim cites a numbered source; a root cause needs a corroborated fact
- Thin evidence means no root cause, and a list of what is missing
- Stored runs by default; live re-runs inside a daily budget
High turnover means the same decision, made hundreds of times a year - and one wrong approval can cost millions.
FMCSA reports that carriers using PSP to screen new hires lowered their crash rate by 8% and their driver out-of-service rate by 17% on average, compared with carriers that did not.
FMCSA Pre-Employment Screening Program$3,604,518 per fatal crash and $91,112 across all police-reported large-truck crashes, in 2005 dollars. FMCSA published an updated crash-cost methodology in 2026; these older figures understate today's costs.
FMCSA - Unit Costs of Medium and Heavy Truck Crashes (Zaloshnja & Miller, 2007)Large truckload carriers have routinely reported turnover around 90% or more on an annualized basis. Much of it is drivers moving between carriers - so the same files are re-screened again and again.
American Trucking Associations- Spotter AI (Sentinel) - MVR monitoring and a driver-safety platform with AI-powered safety scoring and PSP screening.
- SambaSafety - Continuous MVR monitoring and driver risk analytics combining driving records, telematics, CSA and claims data, for employers and insurers.
- Tenstreet - Driver recruiting, onboarding and driver-qualification file management; acquired DriverReach on March 31, 2025.
- J.J. Keller (Encompass) - DOT compliance management, including driver-qualification files.
- Idelic (Descartes) - Safety Suite for driver risk and coaching, now part of Descartes Systems Group.
DQ Ledger is built for the decision layer: a versioned rulebook, a reason for every outcome and the human workflow around it. It makes no claim about how any company above computes its scores, and it is not affiliated with any of them.
Underwrite the sample book, change a rulebook and watch which decisions move, or build a file from scratch.