LOCATION: You must be based in Argentina, Bangladesh, Brazil, Colombia, Egypt, Nigeria, Ghana, India, Pakistan, Indonesia, Turkey, Chile, Peru, Sri Lanka, Saudi Arabia, Nepal, or Mexico. Applicants outside these locations will be automatically rejected.

YEARS OF EXPERIENCE: The JD requires a PhD, postdoctoral experience, or equivalent advanced technical experience in mathematics, statistics, or a closely related field. Your resume must clearly show directly relevant advanced experience or your application will be automatically rejected.

TECHNICAL REQUIREMENT: Strong programming skills in Python, R, Julia, C/C++, Bash, or another relevant language, plus Linux or terminal experience, are mandatory.

We are hiring a Math Expert – Terminal Bench to design realistic terminal-based computational mathematics tasks for AI-agent evaluation.

What you will do

Design authentic multi-step tasks involving numerical analysis, optimization, statistics, mathematical modeling, probability, and differential equations.
Build self-contained terminal environments, datasets, equations, constraints, and expected outputs.
Implement expert solutions and automated tests using appropriate scientific programming tools.
Define rigorous grading criteria covering accuracy, convergence, stability, feasibility, and reproducibility.

Requirements

PhD, postdoctoral experience, or equivalent advanced technical experience in mathematics, statistics, or a closely related discipline.
Strong scientific programming and terminal/Linux skills.
Experience implementing, testing, debugging, and validating computational algorithms.

Engagement and compensation

Five-week contractor assignment.
40 hours per week with 4 hours overlap with PST.
Compensation: $300 per approved task.
Selection includes a mandatory Terminal Bench screening test and one mathematics-domain assessment.

Please apply only if you meet every mandatory requirement above.