The arc
From curiosity
to method
From a young age, Legos were my favorite obsession, providing the freedom to build something sound and symmetrical or tear it apart and start from scratch. As I moved through high school, math came naturally and became my subject of choice. I applied to colleges largely undeclared and landed at Berkeley with little idea of what I wanted to do or who I wanted to be. What drew me was a tension I couldn't truly shake: I was fascinated by human behavior - its patterns, its irrationality, the way real connection actually works - while mathematics kept revealing itself as the most honest way I had to model the world. Reconciling those two impulses left me genuinely curious and two questions kept rising to the surface: how does the mind work and how might technology extend it rather than replace it? My undergraduate degree in Cognitive Science gave me a language for that curiosity - reward modeling, Pavlovian conditioning, belief systems, alongside a grounding in mathematics, computer science, and statistics. A solid foundation, but not yet a clear direction of what to do with it.
I graduated into uncertainty. I recruited and locked in, landing a Software Engineering job at Salesforce. The work was technically solid. I shipped features, maintained coverage, learned what it meant to build things at scale. But that tension I had, the question of how the mind and technology might actually work in sync, never went away.
When I was laid off in the 2023 Salesforce RIFs, I faced a real choice: keep recruiting for jobs where that question would slowly fade or bet on myself to actually pursue it. I chose the bet.
The Master's is where the threads started connecting. Deeply technical Computer Science courses on Neural Networks, High Performance Computing and Classical Machine Learning along with focused coursework in Computational Neuroscience, Human-Computer Interaction and Human Behavior focused ML allowed me to actually build the technical vocabulary to formalize what I had always been interested in. At Georgia Tech Research Institute, it crystallized. I began researching human-AI teaming in high-stakes search-and-rescue and mass casualty response scenarios, looking at how human decision-makers and autonomous agents work together under pressure, how their behavioral patterns co-evolve, and how to build systems that actually model that symbiosis rather than ignore it.
That work is where cognitive science stopped being a background and started being the method.
I am interested in AI systems that genuinely model how human beliefs, preferences, and behavior change over time. Not systems that predict what people will do, but systems that understand why and can adapt to it and work in sync. The technical foundation is there. The question I keep returning to is the same one I started with at Berkeley, now with the tools to actually pursue it.