Remote Claims Operations Expert - AI Trainer ($55-$80 ...
Mercor - Goodyear, AZ
Apply NowJob Description
What you'll do Mercor is partnering with leading AI labs to advance frontier agent evaluations in claims operations. As a Claims Operations Expert, you'll build long-horizon claims tasks that mirror the work you already do, each paired with a deterministic rubric that grades agent performance against verifiable ground truth. Tasks need to have checkable answers; no open-ended essays, no subjective judgment calls. Expect to build scenarios across: FNOL and triage"”intake against required-field checklists, coverage analysis with a documented correct determination, claim assignment against defined routing rules. Adjudication: reserve setting against guidelines with ground-truth reserve amounts, adjuster notes with required elements, settlement or denial letters against policy language Specialized handling: subrogation screening against defined criteria, SIU fraud referrals triggered by rule"‘based red flags, litigation file management against required-document lists. These scenarios will be challenging and take long sessions of focus. Who we're looking for"”3+ years as a claims adjuster, claims operations specialist, or claims supervisor (P&C, workers' comp, or specialty lines) Adjuster licensure in at least one US state strongly preferred Expertise in one or more of the following: a specific claim type (auto, property, general liability, workers comp, medical), coverage analysis and policy interpretation, reserving practice, subrogation, SIU or fraud investigation, a claims system (Guidewire ClaimCenter, Duck Creek, legacy carrier systems) Comfortable reading and producing claims artifacts: FNOLs, coverage opinions, reserve memos, adjuster notes, settlement and denial letters Clear written communication; can articulate reasoning step by step and encode it into deterministic rubrics Located in the United States Compensation $55-$80 / hr depending on domain depth and prior experience. Strong contributors are promoted based on task quality and throughput. #J-18808-Ljbffr
Created: 2026-05-05