The work

You author engineering problems that a strong model cannot bluff its way through, then judge whether its answers actually hold up. A typical task starts with a scenario you construct — a shaft under combined bending and torsion, a shell-and-tube exchanger with a fouling allowance, a bolted joint that has to survive thermal cycling — along with a worked solution, stated assumptions, and the failure modes a careless solver would fall into. You then review model output line by line: is the free-body diagram right, is the correlation valid in that Reynolds range, did it pick a material without checking the service temperature, did it arrive at a plausible number through invalid reasoning.

Most of the value sits in the second half. Marking an answer wrong is easy; explaining precisely where the reasoning broke, which assumption was silently swapped, and what the correct treatment is — in writing another engineer can audit — is the skill AfterQuery is paying for. Expect to cite standards, handbook correlations, or code provisions where they govern, and to be explicit about units and significant figures.

What the screen looks for

  • A real engineering degree and real practice: design, analysis, test, or plant work you can describe in specifics, not job titles.
  • Depth in named subdomains rather than broad familiarity — the screen will pick one thing you claim and push on it.
  • Willingness to be pinned down. Vague hedging on a technical follow-up reads worse than a bounded answer with stated assumptions.
  • Writing that survives review by a second engineer.

PE licensure, an MS or PhD, published work, or standards-committee involvement are stated as pluses, not gates.

Logistics

Fully remote and asynchronous. Work is drawn from a task queue, so hours are yours to set; contributors commonly report 10–20 hours a week, and consistent weekly availability matters more to task allocation than any single large block. Pay is hourly in the $75–115 range as observed on the platform, varying with specialization and task type — treat it as reported, not promised. You will need a machine capable of running whatever CAD, FEA, or spreadsheet tooling your problems require, though many tasks are pencil-and-paper analytical.