The work
You will help improve AI systems that handle insurance claims processes, policy language, and customer communication. The work uses your practical claims experience to test whether AI responses are accurate, relevant, and useful in real claims situations.
- Write questions based on claims investigation, policy coverage, and communication with claimants or other stakeholders.
- Review AI-generated claims responses for factual accuracy, relevance, and practical use.
- Give clear feedback on claims workflows and customer interactions.
- Follow written evaluation guidelines while using professional judgment and discretion.
What it pays and takes
This is independent contractor work that is fully remote within the United States. Assignments are flexible and asynchronous, so you can work alongside other professional commitments.
- Pay: Up to $75 per hour, depending on the assignment.
- Workload: No minimum hour requirement; active assignments may involve approximately 5 to 20 hours per week.
- Location: Open to candidates in the United States.
- Language: Professional fluency in English.
- Experience: At least four years in claims handling or a closely related role.
- Relevant backgrounds include claims specialist, representative, adjuster, examiner, analyst, or claims-focused customer service work.
- You must have hands-on experience investigating claims, interpreting policy coverage, and communicating claims decisions.
- You need strong writing, attention to detail, reliability, independent work habits, and the ability to handle sensitive or proprietary information confidentially.
- Relevant claims experience may include property, casualty, auto, health, disability, or workers' compensation. No prior AI or technical experience is required.
- You need reliable internet access and a desktop or laptop computer.
How it works
Apply on OpenTrain with your resume, then complete the application on the hiring site.
About AI training work
AI training is the human work behind systems that learn from examples, including writing and rating model responses. People with claims expertise help make these systems more accurate by judging whether outputs match real policies, workflows, and customer needs.