Salary: $40-$125/hr
AI-generated summary of the employer’s posting. Always check the full description below before applying.
What They’re Looking For
The role seeks experienced healthcare professionals able to apply clinical judgment and professional standards to realistic medical scenarios. Kyle Howell – [email protected]
Skills to Bring
Clinical Judgment, Clinical Scenario Design, AI Model Evaluation, Clinical Documentation, Clinical Task Design, Clinical Output Grading, Comparative Evaluation, Rubric Development, Differential Diagnosis, Triage, Medication Reconciliation, Drug Interaction Assessment, Red Flag Identification, Clinical Safety Assessment
Medical Expert
| Remote $40-$125/hr USD
**Work when you want**
About the Opportunity
We are seeking experienced healthcare professionals to evaluate and improve next-generation AI models for clinical workflows.
- This is a 100% remote, flexible contract role (not a traditional clinical position).
- Work schedule: Self-directed, minimum 10 hours/week with full-time hours available.
- Location: United States, Canada, or United Kingdom.
Role Overview
You will use your clinical judgment to design realistic medical scenarios (e.g., SOAP notes, triage decisions, medication reconciliations, discharge summaries) and evaluate how well AI models perform against professional standards.
Your main goal is to create authentic tasks that a competent clinician would handle correctly, but an AI model might get wrong, and then clearly document where the AI succeeded or failed.
Key Responsibilities
- Design Clinical Tasks: Create realistic patient cases, prompts, and supporting files (e.g., de-identified histories, lab results, imaging, medication lists) based on real-world practice.
- Evaluate AI Outputs: Run tasks through frontier AI models and grade their responses against professional clinical standards.
- Compare & Rank: Review competing AI outputs on identical tasks, decide which performed better, and document key differences.
- Build Grading Rubrics: Draft detailed rubrics specifying what a correct clinical deliverable must include (e.g., proper differentials, catching red flags, correct drug interactions).
- Identify Failures: Catch and document errors like missed contraindications, unsupported diagnoses, fabricated facts, or unsafe recommendations.