What the work involves
You write respiratory care cases the way they actually unfold — a presenting patient, an assessment, a gas, a set of vent settings — and then document the reference answer, including why the FiO2 comes down before the PEEP or why this particular patient is not a candidate for extubation despite passing an SBT. The clinical reasoning is the deliverable, not just the endpoint. The second half of the work is grading: a model produces a management plan, and you score it for physiologic accuracy and patient safety, then write a short justification another RRT could audit.
The highest-value judgment on this project is catching answers that read fluently and would still hurt someone — permissive hypercapnia applied to a head injury, auto-PEEP mistaken for a compliance problem, a suction depth or cuff pressure that sounds authoritative and is wrong, an ARDS plateau pressure quietly drifting past 30. Frontier labs are paying for exactly this discrimination, because generic reviewers do not have it.
What the platform screens for
- Verifiable credentials: active NBRC RRT and an unencumbered state license, checked directly.
- Bedside depth under follow-up — expect an AI screener to ask a vent or ABG question, then push on the reasoning behind your answer rather than accept the number.
- Ability to articulate reasoning in writing at a level a second therapist could reproduce.
- Realistic availability, since throughput and consistency matter more than raw hours.
Logistics
Fully remote and fully async: no shifts, no scheduled calls, no nights. You set your hours each week against a 10-hour minimum, which most contributors report fitting around clinical work. Pay is hourly via Stripe, disbursed weekly. Project work is ongoing rather than one batch, so consistency week to week tends to matter more than volume in any single week.