What the work involves
This is not a helpdesk seat. Instead of resolving tickets for live users, you produce and judge the reasoning behind resolutions. Typical tasks: take a realistic support scenario — a laptop that gets a DHCP address but can't resolve internal hostnames over VPN, a Windows profile corruption after an update, a Mac that fails 802.1X authentication — and write the full diagnostic path, including what you'd check first, why, and what each result rules out. Other tasks put you in front of a model's answer and ask you to grade it: did it skip layer-1 checks, invent a registry key, confuse DNS suffix search order with split tunnelling, or give a fix that would work but destroy user data?
Good contributors here are the people who, in a real escalation queue, could explain why the fix worked and not just that it did. Bad model answers are often confidently plausible, so the value you add is catching the step that a real technician would never take.
What the platform screens for
- Verifiable Tier 2/3 depth: platforms, tooling, and the kinds of escalations you personally owned.
- Networking fluency under follow-up — TCP/IP, DNS, VPN behaviour explained without hand-waving.
- Cross-OS breadth across Windows, macOS, and Linux rather than a single stack.
- Documentation habit: whether your written steps are reproducible by someone else.
- Willingness to be wrong in writing, with reasoning attached, rather than asserting.
micro1's screening is AI-led and conversational. Expect it to push on one answer three or four times looking for the point where recalled experience runs out.
Logistics
Contractor engagement, fully remote, asynchronous. Observed pay for this listing sits at $30–55/hr, typically tied to demonstrated depth and task type — that band is what has been posted, not a guarantee. Work usually arrives in batches with per-task or weekly deadlines rather than fixed shifts; many contributors run 10–20 hours a week alongside other work. No prior AI or ML experience is expected or asked for.