What the work actually is

You are not processing claims here — you are constructing them. A typical task starts with you picking a messy situation you have genuinely lived: a DRG that shifted on a query response, a secondary payer that recouped after COB was updated late, an auth that was approved for the wrong CPT and denied on appeal for lack of medical necessity. You then build the supporting artifacts — chart excerpts, remit and denial language, payer policy references, portal screens or transaction detail — so that the scenario is solvable but not obvious. After that you either author the expected resolution or evaluate what an AI agent produced against it, writing out where it went wrong and why the error would matter financially or in an audit.

The evaluation half is where domain seniority shows. Agents tend to produce answers that read fluently and are wrong in specific, expensive ways: citing a coding guideline that was superseded, appealing on the wrong basis, missing a timely-filing bar, treating a medical-necessity denial as a coding denial. Your job is to catch that precisely and document it, not to give an impressionistic score.

What the screen is looking for

  • Recency and current employment. Two-plus years hands-on and currently working in a healthcare admin/back-office role. Systems fluency is probed concretely — Availity, Optum/Change, Waystar, Office Ally or payer hubs, EHR/PM platforms, encoders and CAC, clearinghouses.
  • Ownership of exceptions. Appeals you carried end to end, audits you ran, a code assignment you defended, an EHR conversion or payer implementation you led, SOPs you wrote, staff you supervised, committee seats.
  • Breadth. Two or more care settings, or both provider and payer side, is a strong plus. Settings well beyond hospitals count — ASCs, SNF/LTC, home health and hospice, behavioral health, dialysis, infusion, PBM prior auth, DME, FQHCs, health plans, TPAs, RCM outsourcers.
  • Out of scope, firmly: clinical licensure roles, technicians, front desk, registration, scheduling, call-center scripting, scribing, transcription.

Logistics

Fully remote and asynchronous — you pick up authoring and review batches on your own schedule, with a 10+ hrs/week commitment. The US is the primary market; Canada, UK, Ireland, continental Europe, Australia and New Zealand are also open, and non-US candidates are expected to work in their own country's insurance, funding or billing system rather than pretending to US payer knowledge. Pay has been observed in the $50–65/hr range, with placement within it reflecting seniority and priority-area fit; rates are set per engagement and never guaranteed.