Insurance Specialist

What you'll do

This is the most experienced-evaluator seat in Mercor's insurance set. Where the Insurance Underwriting Specialists contract takes 2 years and teaches you to grade, this one requires you to have graded AI output against rubrics already, and adds eight years of top-tier insurance experience on top.

The work, for a leading AI lab's GenAI team:

• Guide research and engineering teams on gaps in underwriting, claims and risk-assessment reasoning
• Design challenging insurance tasks with well-reasoned solutions rooted in real underwriting and claims practice
• Evaluate model outputs against structured rubrics, with written feedback on correctness, judgment and reasoning
• Develop and refine insurance-specific evaluation guidelines and scoring rubrics
• Work with other SMEs to keep training data consistent

The rubric-refinement strand sets this apart from a pure grading job: you shape the criteria as well as apply them.

Who fits

• 8+ years in insurance (underwriting, claims, actuarial or risk management) at a recognised top-tier organisation; the ad's examples are AIG, Chubb, Allstate, Progressive, MetLife and Marsh McLennan
• Mandatory: prior hands-on experience evaluating LLM or AI outputs against rubrics or structured scoring. The ad asks you to describe it in the application, so be specific about the platform, task type and volume
• Demonstrable progression, for example Underwriter to Senior Underwriter to VP of Underwriting
• At least 35 hours a week on weekdays
• United States-based (a hard location requirement)

Marsh McLennan's inclusion means broker-side experience counts, not only carrier roles.

What it pays

$60–80 per hour. With 35+ hours a week that is a substantial income, but for an 8-year, prior-evaluation bar it sits at the same level as the 2-year underwriting and actuarial contracts, which pay a flat $80.

Employment terms

The listing contradicts itself. The body describes a W-2 position with Cincinnatus LLC, the employer of record placing staff at the client lab, while the Mercor footer describes an independent contractor role with weekly Stripe or Wise payments and no H-1B or STEM OPT support. Confirm which applies.

Worth knowing

Good:

• Shapes the rubrics as well as applying them
• Broker and carrier backgrounds both count
• Near-full-time hours offer steady income

Less good:

• US only
• Prior AI evaluation experience is mandatory, which rules out first-timers
• Pay does not rise to match the 8-year bar
• W-2 versus contractor terms are unclear

About this listing

Posted by Mercor as a part-time position for US-based candidates, confirmed open on 24 September 2026. Requirements, hours and pay are the ad's own. See Mercor.