Mercor connects elite creative and technical talent with leading AI research labs. Headquartered in San Francisco, our investors include Benchmark, General Catalyst, Peter Thiel, Adam D'Angelo, Larry Summers, and Jack Dorsey.
- Design challenging computational problems using scientific software to test AI capabilities.
- Develop problems requiring AI to compute exact answers from complex, multi-step workflows.
- Create tasks where AI must strategically plan queries or experiments to uncover hidden information.
- Refine problems through testing loops against state-of-the-art AI models to achieve target difficulty.
- Work independently and asynchronously to refine problem designs based on feedback.
Qualifications
Must-Have
- Graduate-level training in a relevant STEM field (MS, PhD, or equivalent research experience).
- Proven proficiency with at least one of the listed scientific software libraries.
- Solid Python skills for writing problem setups and solution validators.
- Ability to work independently in a Linux/terminal environment.
- Available for at least 15–20 hours per week.
Preferred
- Experience across multiple listed domains or tools.
- Familiarity with benchmark or evaluation design.
- Background in scientific teaching or exam/problem-set design.
- Experience with computational reproducibility and containerized environments.