What the work involves
You are grading and repairing engineering reasoning. A typical task hands you a problem — a bearing selection, a heat exchanger sizing, a fatigue life estimate, a GD&T callout, a tolerance stack-up — along with one or more model responses, and asks you to judge whether the physics, the assumptions, and the arithmetic hold up. Where they don't, you write the corrected solution with the intermediate steps shown, cite the standard or handbook basis (ASME, AISC, Machinery's Handbook, ASHRAE), and explain precisely where the model went wrong. Other task families include authoring original problems at a specified difficulty, writing rubrics for graders who follow you, and ranking competing responses on correctness, unit discipline, and engineering judgment.
The failure modes you'll catch most often are not exotic: unit conversion slips, a moment taken about the wrong point, safety factors applied twice, material properties quoted at the wrong temperature, and confident-sounding answers built on an assumption the problem statement rules out. Reviewers who explain why an approach is wrong — rather than just posting a corrected number — are the ones kept on rotation.
What the platform screens for
- An accredited mechanical engineering degree, and evidence you've practiced: design work, analysis, test, or manufacturing. PE licensure, FE, or graduate work helps but isn't universally required.
- A live technical interview with an AI interviewer that asks follow-ups. Naming a project isn't enough; expect to be pushed on load cases, material choices, and what you would do differently.
- Written clarity. Sample corrections are read closely — sloppy notation, missing units, or hand-waved assumptions end candidacies.
- Domain declarations you can defend. Claiming thermal-fluids and then fumbling a Nusselt correlation is worse than claiming only structural work.
Logistics
Fully remote and asynchronous. Work arrives in batches tied to client projects, so volume fluctuates — some weeks offer 20+ hours, others little. Most contributors treat this as supplemental rather than primary income. Rates are set per project and per specialization; $55–100/hr reflects publicly reported ranges for engineering reviewers on micro1, not a guaranteed band. Deadlines are usually measured in days, not hours, and you choose which tasks to accept.