Mathematics Expert

$80–90/hrrate as listed

micro1 is contracting mathematicians to write and review research-level problems, solutions, and proofs used to train and evaluate frontier AI models. Work is fully remote and project-based, with observed rates in the $80–90/hr range depending on background and task type.

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

What the work involves

You write original mathematics problems at or near research difficulty, produce complete solutions with proofs that survive scrutiny, and review other contributors' submissions for correctness, ambiguity, and unintended shortcuts. A typical batch might ask for problems in a named area — real analysis, measure theory, algebraic topology, PDEs, combinatorics, number theory — with a required difficulty ceiling and an explicit rubric for what counts as a valid solution. You will also grade model outputs: reading a model's attempt at a proof, deciding whether it is correct, and pinpointing precisely where the argument breaks rather than simply marking it wrong.

The reviewing half of the job is where most contributors are separated from each other. Catching that a submitted proof quietly assumes measurability, or that a problem has a two-line trick solution that defeats its intended purpose, is the skill being paid for. Written justification matters as much as the verdict — annotations are read by researchers and by other reviewers, and vague feedback gets rejected.

What the platform screens for

micro1 runs an AI-led interview before any human touchpoint. Expect to state your degree, institution, and advisor or thesis area, then defend a specific claim of depth under follow-up questioning — naming a subfield is not enough, you will be asked to work through or explain something in it. Screens commonly include a live proof or error-detection prompt with no reference materials, and probe whether you can articulate why an argument fails rather than just that it does. LaTeX fluency is assumed. Publications, competition history (Putnam, IMO), or teaching at the proof-based level all help but do not substitute for performance in the screen.

Logistics

Fully remote, contractor engagement, asynchronous with no fixed shifts
Batch-based volume — availability fluctuates with client demand, so treat this as supplemental rather than guaranteed
Most contributors commit 10–20 hours per week; some projects request more during ramp-up
Pay observed at $80–90/hr, varying by specialization and whether the task is authoring or review; rates are set per project and are not guaranteed
Occasional synchronous calibration calls with project leads, usually scheduled around US or European hours

What they look for

Research Methodology
Mathematics
Mathematical proof writing
Advanced Mathematics
Mathematical Problem-Solving

Before you apply

Would you pass this screen?

Drawn from what this listing actually requires. Answered in your browser — nothing is sent to us, and we do not store your answers.

Do you hold a BSc, MSc, or PhD in Mathematics, Applied Mathematics, or a closely related field (e.g., Theoretical Physics, Computer Science with advanced math focus)?
Advanced Mathematics Degree · required
Have you authored or significantly contributed to at least one research paper, published proof, technical analysis, or thesis that demonstrates research-level mathematical rigor?
Research or Proof-Writing Portfolio · required
Are you able to author and verify solutions to advanced calculus problems, including multi-variable calculus, differential equations, or real analysis?
Calculus Mastery · required
Are you able to work remotely and commit a minimum of 10 hours per week to this engagement?
Remote Work Availability · preferred
Is English your working language, or do you have demonstrated fluency in written English suitable for authoring and explaining mathematical content?
Written Communication in English · preferred

0 of 5 answered

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: Whether the candidate has genuinely produced research-level mathematical work and can articulate the depth of reasoning required, not just claim expertise.
A strong answer: Name a concrete publication or documented analysis, describe the central mathematical challenge it addressed, and explain at least one decision made to ensure logical rigor or clarity—such as handling edge cases, selecting a proof technique over alternatives, or structuring the argument for peer review.
Where people lose it: Offering vague references to 'research' or 'advanced problems solved' without naming a specific piece of work, describing its context, or explaining the mathematical reasoning that went into it.