Engineer by training, data scientist by trade, teacher by choice.

I build AI systems that people can trust with real decisions — and I've learned that trust comes from evidence, not from eloquence.

I started my career on factory floors. At Isfahan Steel and later Reef Chemical Industries, I used statistics and machine learning to predict equipment failures. It taught me something I still rely on: a model only matters when the people running the operation trust it enough to act on it.

That question — how do you make a system trustworthy when someone may be actively trying to break it? — took me to the University of Texas at San Antonio. My Ph.D. research used game theory to model attackers and defenders of cyber-physical manufacturing systems and to quantify where security investment pays off. I taught engineering students along the way.

Since 2020 I have been at H-E-B, where I lead AI and data-science work as a hands-on technical lead: demand forecasting that cut forecast error roughly in half, markdown and shelf-space optimization, and, more recently, LangGraph agentic workflows on regulated data. I work directly with executives on what to build, set the evaluation bar for the team, and still write production code.

Outside the day job I build a public portfolio of AI applications for regulated decisions — credit, clinical, trade surveillance — each with independent verifiers, human review and honestly reported limits. I teach agentic and generative AI programs offered by Johns Hopkins University and UT Austin, and I serve as an AI and technology board advisor to the health-tech startup Syne Life.

Outside work

I live in San Antonio, Texas. When I'm not working you'll usually find me climbing, in the mountains, or at a CrossFit box — pursuits that reward the same patience and respect for risk that good engineering does.

Full history on LinkedIn Contact me →