Medical Director, AI Safety

About the role

Own the clinical safety of AI-generated content, setting the bar for what may reach clinicians and verifying it before release. Most of the time goes to safety work, with some module building and customer-facing clinical explanation.

What you will do

• Set the safety bar for each product line, defining hallucinations, omissions and material errors by severity
• Create and run evaluation protocols covering hallucination detection, omission analysis and accuracy against source records
• Maintain a clinical harm taxonomy and severity grading rubric and apply it to benchmark sets
• Review AI-generated clinical content and grade the automated graders
• Own the release gate with go/no-go safety assessments for new modules and major model or prompt changes
• Trace reported clinical errors to root cause and drive fixes, and write validation reports for buyers

What they look for

• Experience judging the clinical safety and quality of AI-generated medical content
• Skill in designing evaluation methods and severity scoring for clinical outputs
• Ability to explain testing and validation to clinical and executive audiences
• Familiarity with the regulatory landscape for clinical AI, including FDA, ONC, CHAI and Joint Commission