Mathematics PhD - AI Evaluation Expert

Mercor; Toronto; On-site; CAD 83,000 - 124,000; Part time

Job summary

Mercor is recruiting PhD and Master's scientists to author AI evaluation tasks for Sci Code, a benchmark project in scientific computing. You will craft original, executable research problems that current frontier models cannot solve, with a focus on at least two deep subdomains and a coding emphasis.

Duration is 6 weeks, part-time at 20+ hours per week, with immediate start. Candidates should have strong Python for scientific computing, experience with GitHub and Docker workflows, and

Qualifications

• PhD in mathematics, applied mathematics, computational mathematics, or closely related field.
• Depth in at least two of: numerical linear algebra, computational mechanics, computational finance.
• Proficiency in Python for scientific computing.
• Experience with Git/GitHub and running code in Docker within PR workflows.

Responsibilities

• Source your own material: published paper, Kaggle dataset, open-source repo, or your own design.
• Write scientific prompts based on input.
• Build grading criteria to define correct answers.
• Calibrate tasks against frontier models; ship tasks when strong models fail.

Skills

Python for scientific computing

GitHub

Docker

Education

PhD in mathematics/applied/computational mathematics or related field

Tools

GitHub

Docker

Job description

Mercor is hiring PhD and Master's scientists to author AI evaluation tasks (Sci Code)

Mercor is partnering with leading AI labs on a new benchmark for scientific computing. You will author original, executable research problems that today's frontier models cannot solve.

Domains - depth required in at least two subdomains (with a coding focus)

• Mathematics — numerical linear algebra, computational mechanics, computational finance

What you'll do

• Source your own material: a published paper, a Kaggle dataset, an open-source repository, or a scenario you design

• Write scientific prompts based on the input

• Build the grading criteria that define a correct answer

• Calibrate against frontier models — a task ships only when strong models fail it more often than they succeed

Required

• PhD in mathematics, applied mathematics, computational mathematics, or a closely related field

• Demonstrated depth in at least two of the following subdomains: numerical linear algebra, computational mechanics, computational finance

• Working proficiency in Python for scientific computing

• Comfortable with Git/GitHub and running code in Docker — authoring runs through a pull-request workflow with automated quality checks

Preferred

• Publications in peer-reviewed journals

• Prior scientific software or research engineering experience

Engagement

• Duration: 6 weeks

• Commitment: part-time, 20+ hours per week

• Start date: immediate