What the work actually involves

You author engineering problems with a defensible correct answer and then judge whether a model got there for the right reasons. A typical item might be a footing sized for a given bearing capacity and load combination, a culvert sized under a specified storm return period, a beam checked against ACI or AISC provisions, or a phasing question on a construction schedule with float implications. Each item needs a full worked solution, the governing code or standard cited by section, stated assumptions, and usually a set of plausible-but-wrong paths that a model might take — wrong load combination, unfactored versus factored loads, imperial/SI slips, a Manning's n pulled from the wrong surface.

The evaluation half is where judgment shows. Models frequently produce answers that are numerically close but reason from the wrong governing case, or cite a code provision that exists but does not apply. Your feedback has to name what went wrong at the engineering level, not just mark it incorrect. Reviewers on these programs also flag your items back to you when a problem is ambiguous or has two defensible answers, so expect revision cycles.

What the screen looks for

  • Real practice history: specific projects, your role, the codes and software you worked under (AASHTO, ACI, ASCE 7, IBC, HEC-RAS, SAP2000, Civil 3D).
  • Depth under follow-up — the screen will push a step past your first answer to see whether you know a method or have memorised a result.
  • Whether you can specify a problem tightly enough that it has one answer, and whether you know when a real design question genuinely has several.
  • Clear written technical English. Every item you produce is read cold by someone who was not in your head.

Logistics

Remote, asynchronous, contract. No fixed shifts; work is drawn from a queue with per-batch deadlines. Most contributors run 10–20 hours a week and are paid hourly against reviewed output. A PE licence, MS/PhD, or standards-committee involvement is preferred rather than required, and typically moves you toward the upper end of the observed band.