What the work actually involves

You will spend most of your time reading model outputs about drugs and judging whether they would survive scrutiny from a clinical pharmacist, a medical reviewer, or a pharmacovigilance assessor. Typical items include a dosing recommendation in renal impairment, a claimed CYP-mediated interaction, a summary of a pivotal trial's efficacy endpoint, an adverse-event causality narrative, or an explanation of a mechanism of action. Your job is to label what is wrong, say precisely why it is wrong, and cite the standard or evidence that settles it — a label with no reasoning attached is usually rejected on QA.

Alongside evaluation, expect authoring tasks: writing reference answers to hard pharmacology prompts, constructing questions a model is likely to fail, and rewriting a flawed response into one you would sign your name to. Some batches ask for side-by-side preference ranking with a written rationale; others are single-response rubric scoring across dimensions like factual accuracy, safety, completeness, and hedging calibration.

What the screen looks for

  • Advanced credential in pharmacy, pharmaceutical sciences, pharmacology, clinical pharmacology, or medicine, and the ability to describe real professional work behind it
  • Genuine depth under follow-up — screeners push past a first answer to see whether you can distinguish a labelled contraindication from a theoretical one, or a surrogate endpoint from a clinical one
  • Calibration: knowing when the literature is genuinely unsettled and saying so rather than inventing certainty
  • Writing that a non-pharmacist annotation reviewer can follow without losing the clinical point

Logistics

Remote, US-based, contractor engagement. Work is asynchronous and pulled from a task queue rather than scheduled — volume fluctuates with project demand, and most contributors treat it as part-time alongside clinical or industry work. Expect an unpaid or lightly paid calibration exercise before live tasks, and periodic recalibration when rubrics change. The $126–441/hr band reflects observed rates across credential tiers and task types, not a guaranteed rate.