What the work actually involves
You receive simulated incident or service requests — text transcripts, call summaries, chat threads, or radio-style exchanges — and do what you'd do on console: determine priority, decide what resource goes, decide what questions get asked first, and decide when it escalates. The difference is that the reasoning is the deliverable. A correct call-type code with no explanation is worth little here; the training value sits in why you chose Priority 1 over Priority 2, which EMD or ProQA-style card you'd open, what caller information would flip your decision, and what you'd hold off on until you had more.
Expect a mix of task types: generating scenarios from your own experience, writing model-facing gold answers, and reviewing AI-produced dispatch reasoning to flag where it invents protocol, skips a caller-safety question, or routes to the wrong agency. You may also be asked to handle multi-channel scenarios where information arrives simultaneously and you must show how you sequence it.
What the screen looks for
micro1's screening is AI-led and follow-up heavy. It is testing whether you have actually sat a console: the interviewer will push on specifics — CAD system names, call-type taxonomies, how your center defined a Priority 1 medical versus a Priority 2, what your QA review process caught you on. Generic answers about "staying calm under pressure" collapse quickly under a second question. It also tests evaluation judgment: can you read a plausible-sounding but wrong AI response and name the specific error rather than saying it "feels off." No AI background is expected or asked for.
Logistics
- Fully remote, asynchronous, contractor engagement.
- Pay is output-based, per accepted task; the $30–55/hr band reflects observed effective rates and is not guaranteed — your throughput drives it.
- A weekly minimum task submission applies; confirm the number before accepting.
- Roles are typically filled within 48 hours and first tasks are expected 24–48 hours after onboarding.