What the work involves

You will be handed AI outputs that look like the work product of a revenue integrity analyst: a flagged missed charge on an OR case, a proposed CDM line change, a duplicate-charge alert, a modifier recommendation, a narrative explaining suspected revenue leakage in an infusion department. Your job is to decide whether each output is defensible against CMS billing guidance, Medicare Part A/B rules, NCCI edits, and payer-specific policy — and then to write the reasoning down in a form a model can learn from. "Wrong" is not enough; the platform wants the citation, the correct alternative, and the compliance exposure created by the error.

Typical task types include scoring paired responses against a rubric, correcting AI-drafted audit findings, writing gold-standard charge review examples from de-identified scenarios, and stress-testing outputs where the model produces plausible-sounding but non-compliant guidance — unbundling, upcoding by implication, or CDM pricing logic that ignores charge-to-cost relationships.

What the screen looks for

  • Verifiable operational history. Five-plus years actually running charge capture, charge integrity, or revenue integrity work — not adjacent RCM exposure. Expect follow-ups on the specific systems you used (Epic Resolute, Cerner, Craneware, nThrive, Optum) and the size of the CDM you owned.
  • Depth under pressure. Questions about revenue code and CPT pairing, charge lag benchmarks, 340B accumulation, and how you documented a corrective action plan that survived audit.
  • Evaluation judgment. Can you separate a factual coding error from a stylistic preference, and can you rank two flawed outputs consistently?
  • Writing. Feedback that a research team can act on, in clear English, without hedging.

Credentials such as CPC, CCS, CHFP, or CRCS help but do not substitute for hands-on audit experience.

Logistics

Remote and contractor-based, typically 10–20 hours per week with some engagements scaling higher during active data collection. Work is mostly asynchronous through Mercor's task platform, with occasional calibration calls with the research team. Pay is reported at $88/hr for this listing; rates on Mercor vary by project, seniority, and assessed quality tier, and engagement length is not guaranteed.