What the work involves
You are assigned to one of two tracks. In authoring, you produce original questions in your subdomain — items that test reasoning a board-certified clinician or doctoral researcher would need, not facts a search engine returns. Each item requires a self-contained stem, one correct option, nine plausible-but-wrong alternatives, a difficulty rating (Medium ≈ intro undergraduate, Hard ≈ advanced undergraduate, Expert ≈ post-graduate and above), a markdown chain-of-thought solution, and 1–5 references from peer-reviewed journals or clinical guidelines. In verification, you audit pre-written items for accuracy, ambiguity, missing information, and solvability, edit where needed, and write a justification for every change.
The hardest part is usually the distractors. Nine options that are each individually defensible-looking — reflecting real diagnostic confusions, dosing errors, causal-inference mistakes, or guideline misreadings — are what make the benchmark discriminating. Items that collapse to a single obvious answer get returned.
What the platform screens for
- Verifiable credentials: MD, DO, PhD, or advanced doctoral candidacy in medicine, biomedical science, or public health. A master's degree can qualify with unusually deep subdomain specialization.
- Depth under follow-up: expect the AI interviewer to press on a specific claim you make and ask you to defend or qualify it.
- Named subdomain: "internal medicine" is weaker than "inpatient heart failure management" or "signal detection in spontaneous adverse-event reporting."
- Citation discipline: familiarity with the primary literature and current guidelines in your area, and the judgment to distinguish a strong source from a convenient one.
- Written clarity: concise, unambiguous English under length pressure.
Logistics
Fully remote and asynchronous, with no fixed shifts. Expected commitment is 10+ hours per week; contributors typically report the first several items taking substantially longer than later ones as the rubric and distractor conventions become familiar. Observed pay for this listing is $94–119/hr, set by credentials, domain scarcity, and track — not guaranteed, and rates can shift between cohorts.