Physician, Medical AI Evaluation

This is a fully remote position, open to applicants in Minnesota.

📋 Description

• Assess clinical EHR vignettes alongside questions, suggested answers, and misleading distractor answers related to diagnosis and treatment tasks.

• Evaluate benchmark-item clinical reasoning quality, focusing on vignette accuracy and completeness, answer transparency, answer correctness and gradeability, as well as trap quality.

• Determine whether reasoning sequences derive answers solely from vignette facts.

• Correct reasoning trails and provide concise rationales.

• Operate within three independent blind reviews followed by physician adjudication.

• Fulfill project responsibilities within the established timeline and acceptance criteria.

• Meet qualifications, engage in a project, complete assigned tasks, and receive compensation.

⛳️ Requirements

• Medical degree (MD or DO).

• Active, unrestricted US medical license, verified during onboarding and re-validated monthly.

• Currently practicing as a physician.

• Board certification or completion of residency training.

• Recent experience in direct patient care; attending-level preferred.

• Extensive, multidisciplinary diagnostic and treatment experience in internal, family, emergency, or hospital medicine.

• Strong clinical reasoning skills, including differential diagnosis, next-step management, and the application of evidence-based guidelines.

• Proficient in written English (C1+).

• Experience in medical AI evaluation, clinical content or exam-question review, or medical annotation/QA is advantageous.

• Not applicable for medical students, unlicensed medical graduates, non-physician clinicians, or non-clinical healthcare roles.

🏝️ Benefits

• Compensation based on approved tasks, approximately $100/hr equivalent; a flat rate for each approved task.

• One-time onboarding bonus of $50, disbursed once your first production task is accepted.

• Additional earning opportunities for high-quality approved tasks.

• Flexible scheduling; decide when and how much to contribute.

• Remote work; collaborate anytime, from any location.

• Project-based involvement tailored to fit around a clinical practice.

• No patient interaction and no malpractice risk.

• Estimated task load of about 10–20 hours per week while the project remains active.