AI Hiring Vendor Annual Impact-Assessment Evidence Ledger for Heads of Talent

Continuous compliance-evidence ledger for Heads of Talent at 500-1500 person US (Colorado-deployer) and EU AI Act high-risk-system deployer tech firms running AI screening/sourcing vendors (HireVue, Eightfold, Paradox). Buyer pastes 5-15 quantified vendor assertions from the active contract and renewal collateral. AE's adversarial council locks each into a falsifiable form. Throughout the year, the talent head enters small-batch resolution events from their ATS: human-reviewer-vs-AI agreement samples (30-100 reqs, 2-4 week cycle), funnel pass-through rates, time-to-hire lift on cohorts, recruiter override frequency. AE grades each locked claim as evidence arrives and accumulates a vendor miss-pattern feeding the annual Colorado AI Act / EU AI Act impact-assessment filing the deployer is legally obligated to produce. Subscription is justified by ongoing regulatory-evidence accumulation per deployed vendor, not vendor-shopping cadence. Demographic parity is tracked as long-horizon signal; primary grading uses claims testable at small N within AE's weekly loop.

AE's falsifiable-claim infrastructure is a near-exact structural fit for legally-mandated AI hiring impact assessments, giving a solo UK operator a defensible £100-300K ARR path on a non-discretionary regulatory clock.

Still in evaluation (phase: ranked). No verdict yet.