Khaled Ebna Abdullah

I’m a licensed physician with 5+ years of active internal medicine practice in Bangladesh, focused on rigorous physician-led clinical reasoning and evidence-based decision-making. Alongside my clinical work, I evaluate AI-generated medical content for accuracy, reasoning soundness, and clinical safety.

In my medical AI evaluation work, I design structured validation frameworks that map evidence retrieval status, assess mapping strength, and trigger safety flags for issues like hallucinations, overgeneralization, missing differentials, and guideline deviations. I’ve built dual-source validation systems that compare AI outputs against authoritative clinical references to ensure evidence alignment.

I also lead comparative model evaluations across multiple LLMs using blind clinical scenario testing to assess reasoning quality, safety-critical gaps, and edge-case handling. From those insights, I implemented practical confidence scoring based on evidence strength and evidence applicability, supporting consistent audit-ready decisions across validation sessions.

My broader ethos is medical quality governance: clear standards, no silent conflicts, transparent escalation for evidence gaps, and education-ready outputs. I’m currently open to remote physician-focused AI evaluation roles where reliability, clinical safety, and evidence integrity are non-negotiable.