What the work involves

The project runs on two tracks and you would carry part of both. The first is clinical data generation: constructing and quality-checking the safety layer of simulated data rooms so that an AI agent working inside them faces realistic material. That means adverse event listings that behave like real listings, SAE records with the messiness of source data, patient narratives that support a causality argument, discontinuation records, and protocol deviations that actually interact with the safety picture rather than sitting inertly beside it.

The second track is pharmacovigilance task authorship. You write full DSURs and PBRERs — and their component sections — to the standard you would submit, then convert that work into evaluation material: the prompt, the golden output, and a rubric granular enough to distinguish a defensible benefit-risk narrative from one that reads well and is wrong. Expect to spend real time on the rubric. Scoring an agent's cumulative exposure reasoning, its handling of signal detection output, or its treatment of investigator versus sponsor causality divergence in a fatal case is where this role earns its band.

What the screen is looking for

Mercor's screening is AI-led and follows up. The named gate is authorship of aggregate safety reports across both pre-commercial (DSUR) and post-commercial (PBRER) reporting — one side only is a common reason candidates stall. Beyond that, screeners probe for active benefit-risk interpretation rather than report assembly: whether you have argued a position against cumulative safety data and exposure estimates, not only compiled sections others signed. Fatal-case workups, including where you landed when sponsor causality departed from the investigator's, come up frequently and reward specific recall.

Logistics

  • Fully remote, US or Canada only.
  • 20–25 hours per week minimum; 30+ preferred. This is not a few-hours-on-the-weekend engagement.
  • Largely asynchronous, with review cycles and calibration against other reviewers' rubrics.
  • 5+ years of experience required, 10–25 preferred; MD, PharmD, PhD, or equivalent.
  • Observed band for this listing is $140–190/hr, typically tracking depth of aggregate reporting authorship. Rates are as observed on the platform, not guaranteed.