I am a Research Engineer at WorkOnward. Before that, I spent several years as a Senior Machine Learning Engineer at Dell Technologies, working on reliability problems in large storage systems.

My research asks one question: how do we build AI systems whose predictions come with honest, measurable guarantees, especially when a wrong answer is expensive?

Most of my academic work is trustworthy machine learning for hardware security. My NSF-funded master's thesis at Cal State Long Beach treated hardware Trojan detection as an uncertainty problem, not a pure classification one, because a detector that is wrong but confident is worse than one that knows when it is unsure. It appeared at ICCAD 2023, DATE 2024, and the Journal of Hardware and Systems Security. My 2023 survey on chaos-based encryption in IEEE Access is my most-cited paper (50 citations). Uncertainty quantification and conformal prediction tie it together, including two COPA 2023 papers and a short practitioner's book. Lately I've moved toward agentic AI and LLM systems: fine-tuning data pipelines, agent communication protocols, and LLM-based vulnerability detection.

I hold 64 granted U.S. patents, mostly from my Dell years, on ML-driven storage reliability such as disk-failure forecasting and backup failure prevention. Several have been cited by later filings from Amazon, IBM, Salesforce, VMware, SAP, Micron, and Databricks (public on Google Patents).