Subject Matter Expert — Clinical AI Quality & Informatics (Generative AI evaluation)
Evaluated Generative AI outputs for clinical accuracy, logical consistency, and adherence to evidence-based medical and public health standards. Designed and performed structured, multi-step assessments to stress-test agentic systems and identify hallucinations, bias, and dataset vulnerabilities. Translated complex healthcare policy and regulatory requirements into actionable quality criteria for AI systems. • Reviewed proprietary healthcare dataset outputs for compliance and safety risks • Provided structured feedback loops to engineering teams to improve ML reliability • Assessed agent behavior against clinical logic and evidence-based guidelines • Identified failure modes such as hallucinations and structural weaknesses