What the work actually looks like

You open a completed item: the task prompt as it was written, a transcript of what the AI agent did, whatever files it produced (a policy draft, a staffing grid, a payer appeal, a schedule, a budget variance memo), and the scoresheet an AI grader already filled in. Your job is review, not production. You answer four things in writing — is this a task that would genuinely come up in this job, is the domain content correct, does the grader's eleven-question yes/no sheet hold up line by line, and does the score match the band you'd assign yourself. Every judgment carries a short written reason, so you're typing all hour, not just clicking.

The judgments that matter most are the ones only a practitioner catches: a HIPAA minimum-necessary framing that's technically wrong, a CMS condition of participation cited from the wrong year, an FTE calculation that doesn't reconcile with the census it claims to support, a payer timeline that no real plan uses. You'll also be asked to disagree with the grader and say why — the platform is looking for reviewers who can hold a position against a plausible-sounding scoresheet.

What the screen looks for

  • Verifiable depth in one of the five named roles — RN, NP, medical or health services manager, first-line supervisor of office/administrative support, or medical secretary/administrative assistant — and enough specificity that follow-ups can go two or three layers deep.
  • Ownership of the work rather than advisory exposure. Chart auditors, preceptors, charge nurses, survey veterans and people who have had to tell a colleague exactly why their documentation failed are the profile.
  • US health care grounding. Tasks turn on HIPAA, CMS and US payer mechanics, so equivalent experience from other systems is a poor fit here.
  • Written English strong enough to justify a call in two or three sentences, and enough spreadsheet or PDF comfort to check whether figures reconcile.

Logistics

Fully remote and asynchronous, paid hourly in an observed $75–90/hr band (as reported for this project, not guaranteed). Items run about an hour each and you take them as your schedule allows; there is no shift coverage. Entry is a roughly 15-minute recorded interview about your own area of expertise, and your first task on the project is a short intake profile you complete unaided — treat both as evaluated work.