Insurance Expert for AI Model Evaluation & Training

SaidGig; New York (NY); On-site; USD 83,000 - 110,000; Full time

Job summary

SaidGig in the United States is seeking an experienced insurance professional to help develop advanced large language models by applying real-world underwriting, claims, and risk assessment judgments to AI training and evaluation.

You will collaborate with research and engineering teams, create tough insurance-focused tasks, evaluate model outputs with rubrics, and help refine scoring standards while ensuring high-quality training data.

Qualifications

• At least 8 years of dedicated professional insurance experience in underwriting, claims, actuarial work, or risk management at a recognized top-tier organization.
• Hands-on experience evaluating LLM or AI model outputs against rubrics or structured scoring criteria.
• Career progression such as advancing from Underwriter to Senior Underwriter to VP of Underwriting.
• Strong verbal and written communication, problem-solving, and interpersonal skills.

Responsibilities

• Partner with research and engineering teams to address knowledge gaps in underwriting, claims, and risk-assessment reasoning.
• Create challenging insurance-focused tasks and accurate, well-reasoned solutions grounded in practical underwriting and claims experience.
• Evaluate AI model responses using structured rubrics and provide clear written feedback on accuracy, judgment, and reasoning quality.
• Develop and improve evaluation guidelines and scoring rubrics for insurance tasks.
• Work with fellow subject-matter experts to maintain consistent, accurate training data.

Skills

Underwriting

Claims experience

Risk management

Communication skills

Job description

SaidGig in the United States is seeking an experienced insurance professional to help develop advanced large language models by applying real-world underwriting, claims, and risk assessment judgments to AI training and evaluation.

You will collaborate with research and engineering teams, create tough insurance-focused tasks, evaluate model outputs with rubrics, and help refine scoring standards while ensuring high-quality training data.