Policy, Coverage & Regulatory Compliance Expert for AI Research Project

About Aptura

Aptura works with leading foundational AI labs to bring institutional finance expertise directly into AI model development. Founded by ex-Lazard and Partners Group professionals, we operate from London and San Francisco.

The Opportunity

You’ve built deep expertise in insurance coverage, policy wording, product development, or regulatory compliance, and produced work that drives real decisions. Now that expertise has a new application: shaping how frontier AI models reason through complex insurance problems.

What you'll do

• Design realistic policy, coverage, and compliance tasks: policy interpretation, endorsement analysis, coverage determinations, form reviews, and regulatory compliance assessments without clean answers.

• Write evaluation rubrics that reflect how experienced coverage and compliance professionals reason through policy language, exclusions, endorsements, regulatory requirements, and filing standards.

• Review AI-generated coverage analyses, policy interpretations, compliance assessments, and form recommendations, identifying where models misread policy language, overlook endorsements, or propose non-compliant solutions.

• Build reference coverage opinions, policy review materials, and compliance frameworks that set the benchmark models for testing.

• Communicate coverage determinations, compliance assessments, and rationale clearly in writing.
  

How Engagements Work

Fully remote, part-time, and flexible. Work is scoped by project and ranges from short engagements to ongoing collaborations, structured as a contract or a monthly retainer.

You May Be A Good Fit If:

• 5–8+ years in insurance coverage, policy wording, product development, forms management, or regulatory compliance.

• You’ve drafted, reviewed, interpreted, or approved policy language, endorsements, forms, or regulatory filings that had to withstand legal, regulatory, or commercial scrutiny.

• CPCU, ARM, JD, ACII / FCII, or equivalent (preferred, not required).

• You can explain why a policy responds — or does not respond — clearly enough to teach the reasoning, not just the outcome.

Compensation