What the work actually looks like
You will not be doing production design work for a client. You'll be producing the reference material an AI system learns from: writing realistic mechanical design prompts, working problems the way a practicing engineer would, and grading model output against what a competent designer would actually accept. Typical tasks include critiquing an engineering drawing for datum scheme errors and unmanufacturable callouts, running or checking a worst-case and RSS tolerance stack, judging whether a proposed feature can be molded, cast, or machined as drawn, and explaining the reasoning behind a modeling approach — feature order, sketch constraints, assembly mating strategy — in prose clear enough to train on. Some tasks are open-ended authoring; others are side-by-side comparisons where you rank two model responses and write the rationale.
What the screen is looking for
micro1's process is an AI-led interview followed by task calibration. It probes depth: expect follow-ups on ASME Y14.5 specifics, position vs. profile tolerancing, bonus and datum shift, material selection under wear or contact loading, and how you'd handle a stack that closes on paper but fails in the shop. Vague answers get pressed. The other half of the assessment is evaluation judgment — whether you can separate a confident-sounding wrong answer from a correct one, articulate why a model's GD&T callout is invalid rather than just flagging it, and write feedback a non-engineer reviewer can audit. Communication carries real weight here; unclear rationales are the most common reason otherwise-qualified engineers wash out.
Logistics
- Fully remote, asynchronous, contractor engagement — no fixed shifts.
- Compensation is output-based: paid per task that meets spec, not per hour logged. The $66–130/hr band reflects observed effective rates and depends on your throughput and task tier; it is not a guaranteed wage.
- A weekly minimum task submission applies. Treat this as a consistent part-time commitment rather than occasional pickup work.
- Hiring moves fast — roles typically fill within 48 hours, with first tasks expected 24–48 hours after onboarding.
- No prior AI or ML experience is required or expected.