Numerical Analyst - AI Trainer

DataAnnotation; Michigan; Remote; USD 55,000 - 172,000; Full time

Benefits offered by this job

Flexible, remote projects

Job summary

DataAnnotation is seeking a Mathematician to evaluate and correct mathematical reasoning across everything from competition problems to graduate-level material, setting the standard the models learn from.

You’ll craft problems exposing weak reasoning, review AI solutions step by step, and write gold‑standard solutions with the rigor of a referee report. Flexible, remote, and project-based, with competitive hourly pay.

Qualifications

• Strong background in pure or applied math through graduate work or equivalent.
• Ability to write precise, well-structured proofs in English.
• LaTeX familiarity is a plus.

Responsibilities

• Craft problems exposing weak mathematical reasoning.
• Review AI solutions step by step for invalid derivations and unjustified leaps.
• Write the gold-standard solution with rigorous referee-level critique.

Skills

Strong background in math

Proof writing

English clarity

LaTeX familiarity

Education

Graduate study in mathematics

Competition math

Tools

LaTeX

Job description

Frontier models still fail at rigorous multi-step mathematics: invalid derivations, unjustified leaps, and confidently wrong final answers. Automated checks can't grade a proof. You can.

As a Mathematician you'll evaluate and correct mathematical reasoning across everything from competition problems to graduate-level material, and your judgment becomes the standard the models learn from.

What you’ll actually do

• Craft problems that expose weak mathematical reasoning, from olympiad-style puzzles to graduate-level proofs.
• Review AI solutions step by step: catch invalid derivations, flag unjustified leaps, and rate the quality of the reasoning.
• Write the gold-standard solution when the model falls short, with the rigor of a referee report.

What we look for

• Strong background in pure or applied mathematics through graduate study, competition math, or equivalent experience.
• Fluency writing precise, well-structured proofs (LaTeX comfort helps).
• Clear written English: your explanations are the training signal.
• No degree required. We care about what you can do, not where you learned it.

Compensation

Up to $40 - $125+/hr depending on task difficulty and specialization. Many contributors add $10k-$100k+ a year; some make it their full-time income.

About DataAnnotation

DataAnnotation is where 100k+ experts train the world’s leading AI models. $150M+ paid to contributors to date, and the average contributor stays 5+ years. Flexible, remote, and always project-available.