$94–119/hr · Mercor · Hourly
Medical expert creating and verifying rigorous academic questions for AI evaluation across health science domains.
What you would do
- Create original medical questions testing deep conceptual understanding with precise, self-contained problem statements
- Rate question difficulty as Medium, Hard or Expert based on expected solver background
- Develop one correct answer and nine plausible distractors reflecting common misconceptions and challenging mastery
- Articulate step-by-step reasoning solutions with clear logic explaining why the correct answer is sound
- Supply 1-5 peer-reviewed academic references per question from established clinical and research sources
Who they want
- MD, DO, or PhD qualification in Medicine, Biomedical Sciences, Public Health or closely aligned field
- Mastery of post-graduate medical concepts, reasoning processes, and biomedical research methodology
- Master's degree may qualify if you bring exceptional expertise in a particular medical subdomain
- Board certification, hands-on clinical experience, or research contributions in health sciences strongly preferred
- Excellent written English and capacity to express intricate ideas with clarity and precision; minimum 10 hours per week commitment
Main skills
What the interview asks about
1.Creating challenging medical questions
Well-designed questions distinguish genuine mastery from test-taking skill, critical for validating AI reasoning about medicine.
For example: “Write a question on acute coronary syndrome diagnosis that tests whether a clinician understands the relationship between troponin timing, ECG changes and risk stratification. How would you challenge an expert solver?”
2.Designing subtle distractors
Plausible wrong answers that trip up knowledgeable but imprecise reasoners reveal gaps in clinical logic.
For example: “You're writing about a drug-drug interaction. Create three distractors based on pharmacological misconceptions a physician might hold: one about enzyme induction timing, one about protein binding, one about renal clearance.”
3.Articulating step-by-step reasoning
Clear explanations of clinical logic enable AI systems to learn sound medical judgment, not just pattern-match to answer choices.
For example: “A patient presents with hypokalemia and the correct answer is 'hypomagnesemia prevents potassium repletion.' Write a 3-4 step explanation of the cellular mechanism backing this answer.”
4.Identifying question flaws in verification work
Catching ambiguity, incomplete information or unsolvability preserves question quality and demonstrates critical domain knowledge.
For example: “A question asks about drug dosing in renal failure but doesn't specify creatinine clearance category. How would you revise this and what additional clinical context is needed?”
A task you may get
Author a Medium and an Expert question in your specialty with correct answer, nine distractors, reasoning and references. Review a sample question and identify 3 quality issues.
How to prepare
- Review 2-3 published medical board exam or continuing education questions to study formatting and difficulty calibration
- Identify a specific clinical scenario you know deeply and map out the conceptual nodes a well-designed question could probe
- Compile a personal reference library of 10-15 key papers or guidelines in your specialty area
- Draft a rubric distinguishing a superficial versus rigorous chain-of-thought explanation for a complex clinical decision
The facts
- Pay
- $94–119/hr
- Hours
- Hourly
- Where
- Remote
- Field
- Medicine
- Posted
- 8/13/2026
- Places left
- 5
We wrote this page from the public Mercor listing. It may be out of date, so read the full posting before you apply.