What the work involves

The research team behind this pilot wants to know whether AI-generated analysis of drug assets survives contact with someone who has underwritten these names professionally. Your output is two-sided: you write and refine the rubrics that define what a good asset assessment looks like, and you apply those rubrics to a working set of about 5–10 companies. Expect to move between mechanism-of-action plausibility, trial design and readout quality, competitive and regulatory positioning, and the commercial assumptions that carry a valuation.

A large share of the time goes to thesis critique. You will read investment-style write-ups and mark where the reasoning is thin: an efficacy delta treated as durable on a single Phase 2, a peak-sales estimate that ignores an entrenched standard of care, a safety signal waved through, an approval timeline that assumes a clean panel. Written justification matters more than the verdict — the team is training on your reasoning, not your score.

What the platform screens for

  • Specialist-fund background. Mercor's screen probes where you sat, what you covered, and what you actually underwrote. Generalist healthcare coverage or broad life-sciences consulting typically does not clear this bar.
  • Depth under follow-up. Expect the interview to pick one therapeutic area you name and keep going: endpoints, comparators, the specific readout that moved your view.
  • Calibration. Screens favor analysts who separate what the data supports from what the narrative implies, and who can say where they were wrong before.
  • Written clarity. Rubric work is writing work; terse, defensible prose beats volume.

Logistics

Remote and US-based, asynchronous, roughly 10–20 hours across a 1–2 week window. Pay in this category has been observed in the $120–200/hr range depending on background and pilot scope — treat that as observed, not guaranteed. Strong contributors are typically invited back as the program scales, but the initial commitment is the pilot only.