Senior AI Product Manager focused on evaluating, monitoring, and hardening machine learning systems in production.

I build tools for:

• Medical AI model monitoring (real-world drift detection, calibration analysis, post-market evaluation)
• Statistical evaluation frameworks for calibration, threshold selection, and robust performance analysis
• LLM behavioral reliability (drift, safety boundaries, consistency) — in progress

My work sits at the intersection of research and product, with an emphasis on making AI systems measurable, predictable, and operationally safe.

• Model Eval & Drift Lab
  Tools for detecting distribution shift, assessing calibration, and stress-testing deployed ML systems.

• GPT-Drift
  Lightweight behavioral fingerprinting to detect silent behavior changes in LLM APIs.