AI Trainer (Mathematics & Statistics)

$60–120/hr

rate as listed

Workload: 15+ hrs/week
Engagement: Project-based
Location: Worldwide
Field: Generalist

Create advanced mathematics and statistics tasks, reference solutions, and evaluation criteria for AI systems. This contractor role requires prior paid AI-training or model-evaluation experience, strong Python proficiency, and documented depth in a relevant mathematical specialization.

The work

You will design realistic computational mathematics, statistics, or optimization tasks for advanced AI systems. Day to day, that can include writing precise instructions, building reproducible environments, producing reference solutions, defining objective grading criteria, and reviewing model-generated proofs, calculations, code, models, and optimization outputs.

What micro1 screens for

The AI interview includes technical and integrity checks. Expect follow-up on specific paid AI-training or model-evaluation projects, your mathematical specialization, and examples of public technical evidence. Python is required. Experience with NumPy, SciPy, SymPy, pandas, Pyomo, OR-Tools, Lean, Coq, or comparable tools is preferred, as is experience creating tasks, rubrics, graders, reference answers, or benchmark problems.

Evaluation judgment

Strong candidates can distinguish a plausible answer from a correct and reproducible one. The screen may test whether you can identify invalid assumptions, logical gaps, numerical instability, inappropriate statistical methods, infeasible optimization results, unreliable code, or formal proofs that do not establish the stated claim.

Logistics

This is a global, fully remote contractor role with a flexible schedule and a stated commitment of at least 15 hours per week. Work is task-based, payment depends on accepted tasks meeting project specifications, and task availability and completion time may vary. Immediate availability is preferred.

Pay band: $60–120/hr

What they look for

Mathematics & Statistics
AI Training
Programming Skills

Included with this listing

Interview prep — 10 questions

Matched to what this platform's screening actually measures. Answer each out loud in 60–90 seconds: claim → concrete example → implication.

What it tests: This tests whether your experience involved substantive model evaluation rather than ordinary data analysis, tutoring, or software development labeled as AI work.
A strong answer: Name the project context, mathematical domain, your exact responsibility, the artifacts you produced, and how submissions were accepted or rejected. Include a concrete example of a model failure you identified and how you verified it.
Where people lose it: Candidates lose points when they describe a confidential project only in broad terms and cannot state what they personally evaluated, created, or measured.

Summary, eligibility guidance and screening preparation are prepared by Sidequest from the platform's posting. Check the source listing before applying.