What the work actually involves

You will not be designing hardware here — you will be judging engineering reasoning about hardware. Day to day that means reading a model-generated or human-drafted artifact (a requirements decomposition, an interface control document, a verification matrix, a test procedure, an FMECA, a root-cause writeup, a production-test flow) and deciding whether it would survive contact with a real program. Where it wouldn't, you explain why in terms an engineer would recognize: the missing environmental margin, the untestable requirement, the fault tree that skips a credible interface failure, the test that measures the symptom and not the mechanism.

The other half is authorship. Tasks often ask you to construct hard, well-posed problems from your own domain — a bring-up anomaly with enough instrumentation detail to be diagnosable, a qualification plan against a specific standard, a yield excursion with plausible defect signatures — and to write the reference solution that defines what "correct" means. Good problems here are ones where a confident but shallow answer is clearly distinguishable from a correct one.

What the screen is looking for

  • Recency and hands-on depth. Mercor's screen probes for at least three recent years where you were personally in the lab, on the line, or in the failure-analysis loop — not approving other people's reports.
  • Ownership of a lifecycle slice. Being the responsible engineer for a subsystem through integration, qualification, and field support carries far more weight than broad exposure.
  • Cross-domain reasoning. Most interesting hardware failures live at boundaries: firmware timing against power sequencing, thermal against mechanical tolerance stack, DFT against production takt time.
  • Calibrated critique. Can you separate "wrong" from "unconventional but defensible," and say which is which without hedging everything.

Logistics

Fully remote and asynchronous. Engagements are typically part-time and flexible — many contributors work 10–20 hours a week alongside a full-time role, though availability expectations vary by project and higher weekly commitment often means more task flow. Observed pay for this listing is $100–120/hr; bands on Mercor depend on assessed depth and the specific project, and are not guaranteed. Expect a screening interview, sometimes a written or recorded technical task, and calibration feedback on your first batch of reviews. Some projects ask about export-control status or prior employer confidentiality, so be ready to describe past work at a level you're free to share.