What the work actually looks like

You are producing training and evaluation data for coding models, and the raw material is real engineering work. A task might be: implement a user-facing feature plus the API and data model behind it; drop into a repository you have never seen and trace a failing request from the browser through the service layer to the database; port a working build from one language to another and prove the behaviour survived the translation. Other tasks run the other direction — a model has produced a solution and you judge whether it is correct, complete, performant, secure, and something a maintainer would accept in review. The failure modes you are expected to diagnose are named explicitly in the brief: application state, API contracts, authentication, persistence, networking, deployment.

What the screen is looking for

micro1's pipeline is screening questions, a roughly 30-minute AI interview, an optional technical assessment, then hiring-manager review. The interview probes for genuine both-sides experience — front-end specialists who gesture vaguely at "the backend team handled that" tend not to survive follow-ups, and the same is true in reverse. Expect pressure on production experience in more than one language, on how you orient inside an unfamiliar codebase, and on the judgment side: can you articulate why an AI-generated solution that passes the tests is still wrong? Concrete systems, concrete bugs, concrete decisions beat framework name-dropping.

Logistics and pay mechanics

  • Fully remote, global, contractor status, around 15 hours per week.
  • You pick your hours and days, weekends included, provided there is consistent activity.
  • Compensation is output-based: you are paid per task that meets project specifications, with a weekly minimum submission requirement. The $50–100/hr figure is an observed band that reflects how quickly experienced people clear tasks — it is not a guaranteed hourly rate.
  • Roles typically fill within 48 hours, and selected candidates are expected to start tasks within 24–48 hours of onboarding. If you cannot begin nearly immediately, say so rather than stalling.