Records Manager for AI Model Evaluation

Records Managers apply archival, library, and collections expertise to evaluate and improve AI-generated content related to records, archives, and information management. In this role you will create prompts that reflect real workplace tasks, review model outputs for accuracy and relevance, and provide clear, structured feedback that helps models better understand professional language and workflows.

• Develop prompts that reflect records management, archival practice, cataloguing, reference, and collections workflows.
• Review and evaluate AI-generated responses for factual accuracy, completeness, appropriate tone, and alignment with professional standards.
• Assess content related to your field of work, including catalog records, finding aids, accessioning notes, provenance statements, and reference answers.
• Deliver clear, structured feedback and annotations that explain errors, ambiguities, or improvements to model outputs.
• Work asynchronously to meet project deadlines and quality standards, learning new evaluation methods as required.

• At least 4 years of professional experience in one or more of the following areas: archives, records management, library services, collections management, or information management.
• Relevant real-world responsibilities that provide context for evaluating AI, such as:
• Managing and organizing archival records, developing classification systems, and maintaining databases to ensure access and retrieval.
• Providing reference services, authenticating and appraising historical documents, and coordinating educational or outreach programs.
• Directing staff in cataloguing and preserving collections, researching and acquiring materials, and maintaining preservation standards.
• No prior AI development experience is required; domain expertise and professional judgement are the primary qualifications.
• Ability to participate in asynchronous work with external AI research teams.

• Location: Remote.
• Employment type: Part time.
• Schedule: Asynchronous and flexible, no minimum weekly commitment; most contributors average 5 to 20 hours per week.
• Typical tasks: developing field-specific prompts and evaluating model responses independently.
• Project placement depends on availability. The program runs year-round and new project opportunities open periodically across different subject areas.

• Pay: Up to $75/hr (depending on the project).

• F-1 students who are eligible for CPT or OPT may be eligible to participate; consult your Designated School Official to confirm whether a specific project meets your school’s CPT requirements. If your school requires that a CPT enrollment be tied to a course, these projects may not meet that requirement.
• STEM OPT is not supported for participation in these projects.

• Create an account on the program portal and complete the basic profile information.
• Verify your identity as part of onboarding.
• Indicate your areas of expertise and join the onboarding flow for relevant projects.
• Once onboarded to a project, begin asynchronous tasks and submit evaluations to start earning compensation.