What the work involves

You'll be handed AI outputs drawn from realistic revenue cycle scenarios — an 835 file parsed into posting transactions, an EOB interpreted into line-level allowances and patient responsibility, a lockbox batch reconciled against a bank deposit — and asked to judge whether the model got it right. That means checking whether CARC/RARC codes were applied to the correct service lines, whether a takeback or offset was handled as a reversal rather than a new adjustment, whether contractual write-offs were split correctly from denials, and whether unapplied cash was flagged instead of force-posted. Where the output is wrong, you write the correction and the reasoning behind it: not just "this is misposted" but which payer methodology or remittance convention the model missed.

A second stream of work is authoring scenarios. Labs need cases that break naive automation — split remits across two deposits, a payer recoupment netted against unrelated claims, a secondary payer posting where the primary allowed amount is inconsistent, an EOB with a bundled payment across multiple CPT lines. Managers who have built posting SOPs and variance controls tend to produce the most useful material here.

What the screen looks for

Mercor's process is AI-led: a recorded conversational interview, resume verification, and often a short work sample. The interviewer probes depth rather than breadth — expect follow-ups on specific adjustment reason codes, how you'd trace a reconciliation variance, and what your unapplied cash aging thresholds were. Vague process descriptions get pressed. Reviewers also assess written feedback quality, since annotation rationale is the actual deliverable; if you can't explain a posting error clearly in two sentences, the work doesn't transfer.

Logistics

  • Fully remote, contractor engagement, async in practice with occasional calibration calls
  • Typical commitments range from 10–20 hours weekly; some projects scale up during active data collection windows
  • Work is task-batched, so hours are self-scheduled against deadlines rather than fixed shifts
  • Pay bands observed in this category cluster near $85/hr but vary by project, seniority, and whether you're reviewing or authoring