Remote | Senior Insurance & Underwriting Evaluation Specialist — $55–$75/hour

24-Mag•New York, NY

2d•Remote

About The Position

This is a specialized full-time consulting opportunity for US-based insurance professionals with deep experience in underwriting, claims, actuarial analysis, risk management, and structured evaluation of AI-generated outputs. The role supports a high-impact generative AI initiative focused on improving how advanced models reason through real-world insurance scenarios. Selected professionals will design rigorous domain-specific tasks, evaluate model outputs against structured criteria, develop scoring frameworks, and provide practical feedback grounded in senior-level underwriting, claims, and risk-assessment experience.

Requirements

At least 8 years of dedicated professional experience in insurance.
Senior-level expertise in underwriting, claims, actuarial work, risk management, or a related area.
Experience working within a recognised insurer, broker, consultancy, financial-services organisation, or comparable institution.
Prior hands-on experience evaluating LLM or AI-generated outputs using structured rubrics or scoring criteria.
Demonstrable career progression into senior, management, director, vice president, or comparable leadership responsibilities.
Strong professional judgment and the ability to assess complex insurance decisions.
Excellent written and verbal communication skills.
Reliable availability for at least 35 hours per week during weekdays.
Prior experience evaluating AI or LLM outputs against structured rubrics is required.
Immediate availability is preferred.

Nice To Haves

Experience across property and casualty, life, health, commercial, specialty, or reinsurance markets.
Background in complex underwriting, claims investigation, pricing, reserving, or portfolio risk.
Experience creating structured scoring rubrics, evaluation guidelines, or quality frameworks.
Familiarity with model training, human-feedback workflows, annotation, or AI quality assurance.
Strong understanding of policy language, coverage interpretation, risk selection, and claims decision-making.
Experience reviewing underwriting files, claims assessments, actuarial analyses, or risk reports.
Previous collaboration with product, analytics, legal, compliance, research, or technical teams.

Responsibilities

Guide research and engineering teams on underwriting, claims, actuarial analysis, policy interpretation, and risk-assessment concepts.
Identify gaps in model understanding across coverage decisions, loss evaluation, pricing, claims handling, and insurance operations.
Apply practical insurance judgment to complex scenarios involving policyholders, exposures, financial risk, and regulatory considerations.
Ensure insurance tasks reflect realistic professional decisions and industry standards.
Design challenging, domain-relevant tasks grounded in real underwriting, claims, actuarial, or risk-management practice.
Write accurate, well-reasoned solutions covering insurance analysis and decision-making scenarios.
Develop problems requiring technical judgment, risk evaluation, policy interpretation, and assessment of competing considerations.
Ensure tasks are clear, internally consistent, and suitable for structured evaluation.
Evaluate AI-generated insurance responses against established rubrics and scoring criteria.
Assess correctness, professional judgment, reasoning quality, relevance, and practical applicability.
Compare alternative outputs and determine which response provides the stronger insurance analysis.
Identify factual errors, unsupported assumptions, weak risk reasoning, and incomplete conclusions.
Provide clear written feedback that supports improvements in model behaviour.
Develop and refine evaluation guidelines for underwriting, claims, actuarial, and risk-management tasks.
Define scoring criteria covering technical accuracy, professional judgment, reasoning, and decision quality.
Participate in calibration activities with other insurance specialists.
Collaborate across teams to maintain consistency and accuracy throughout the training data.