Government Quantitative Professional for AI Model Evaluation
Up to $120/hr (depending on the project)
RemotePart-timetechnology
Apply NowAbout this role
Government quantitative professionals apply their mathematical and data science expertise to evaluate AI-generated work, create realistic field-specific prompts, and provide structured feedback that improves how models understand government-sector tasks and language.
Key Responsibilities- Assess AI model outputs and content related to your professional field.
- Develop prompts that reflect real government quantitative work, then evaluate the responses.
- Provide clear, structured feedback to strengthen model performance.
- Draw on experience such as mathematical research, model development, numerical analysis, computational methods, and practical problem-solving across science, engineering, business, or security.
- Contribute perspective informed by mentoring, publishing or presenting research, and ongoing professional learning.
- At least 4 years of professional experience in a government-sector quantitative role, such as mathematician or data scientist.
- Ability to work independently and collaborate asynchronously with leading AI labs.
- Remote, part-time, temporary, and asynchronous work.
- Flexible scheduling with no minimum commitment; typical participation averages 5-20 hours.
- Projects are available year-round, with opportunities opening periodically across areas of expertise.
- Project placement depends on current availability. If a suitable project is not available when you apply, additional opportunities may open later.
Up to $120/hr (depending on the project).
Eligibility- F-1 students eligible for CPT or OPT may be eligible for projects. Consult your Designated School Official to confirm eligibility.
- If your school requires a CPT course, the work may not satisfy that requirement.
- STEM OPT is not supported.
- Create an account and complete identity verification.
- Complete onboarding for relevant available projects.
- Begin contributing once placed on a project.