Opportunity Snapshot

Pay: $45 - $90/hour.

Commitment: Flexible — fits alongside your current role

You’ll need: A completed PhD in a STEM field, plus hands-on research experience

Your tasks: Build difficult technical problems from your own research and define what makes an answer correct

These tasks are hard to build, and that’s the point.

• Design difficult technical problems drawn from your own research area, with realistic constraints and supporting data.
• Provide the correct solution and walk through the reasoning that gets there.
• Define objective criteria for correctness, so another researcher reviewing the work would reach the same verdict.
• Review problems built by other contributors for technical accuracy and soundness of method.
• Flag problems that turn out to be ambiguous or ungradable rather than submitting them.

What we’re looking for

• Completed PhD in a STEM field. This is a hard requirement.
• Hands-on research experience — work you did yourself, not work you read about.
• Wet lab, research lab, or academic experience is strongly preferred.
• Able to explain why an answer is correct clearly enough that another researcher could verify it.
• Publications or a record of research contribution is a strong plus.
• Prior experience on AI training or evaluation projects is a plus, not a requirement.

How the selection process works

1. Apply and complete the screening questions.
2. Complete the AI interview for this role — it’s how your expertise gets in front of the client.
3. Profile review.
4. Onboarding — our Talent Ops team will get you set up.

Work arrangement: Fully remote

Eligible locations: Worldwide