What the work involves

You author scenarios that look like a real pharmacy queue: a handwritten-style sig that could be read two ways, a Rx quantity that doesn't reconcile with the directions, a rejected claim with a DUR code that needs resolution before dispensing. For each one you write a reference answer that shows the correct order of operations — verify, clarify, calculate, adjudicate, flag — not just the final label. The grading side is the mirror image: you read model output and decide whether it would survive tech verification and pharmacist check, or whether it produced a plausible-looking answer built on a misread sig, a wrong days supply, or a missed interaction.

Most tasks are small and self-contained. Typical work items include:

  • Writing a prescription-processing scenario with a specific failure mode embedded
  • Documenting the correct workflow steps with the reasoning at each one
  • Scoring model responses against a rubric for dosing math, sig accuracy, and safety flags
  • Writing a short justification for each score that a reviewer can audit

What the platform screens for

AfterQuery's screen is AI-led and follow-up heavy. It checks that your certification and practice history are real and current, then probes depth: convert a sig, compute days supply on a tapering or PRN prescription, explain what a given rejection code means and what the tech does next. Expect it to push a second and third layer on any answer you give — vague answers get re-asked rather than accepted. It also tests evaluation judgment, which is distinct from pharmacy skill: whether you can tell a model answer that is wrong from one that is merely written differently than you would write it, and whether you can articulate a scoring decision in writing.

Logistics

100% remote, fully asynchronous, no scheduled shifts or calls. You set your hours each week against a 10-hour minimum; pay is hourly via Stripe on a weekly cycle. Work is ongoing project-based rather than a fixed contract term, and volume varies with lab demand. You need a computer and reliable internet — no pharmacy software licence or institutional access is required, since scenarios are authored from your own knowledge rather than pulled from live patient records. Do not bring real patient data into any task.