It started with toys, moved on to drones, and had turned into robotics by the time I started my undergraduate degree. There was never a career plan behind any of it. I just wanted to know how the thing worked, and the fastest way to find out was usually to open it.
Most of what I built in college aimed at problems I could see from where I was standing: low-cost robotic prosthetics, traffic signals that could clear a path for an ambulance, a spatiotemporal model for forecasting air quality. Covid and a complete absence of funding ended nearly all of it. What stayed with me was the pattern underneath. Every one of those projects lived or died on data: data to find the problem, data to study it, data to know whether the fix actually worked.
Which is a slightly unromantic conclusion for someone who wanted to build robots. It is also what sent me to USC for a Master's, where I specialized in data at scale and in the less glamorous half of it: governance, and using it carefully.
The career took the scenic route from there. Business analyst, data engineer, AI engineer, now lead engineer and architect, across finance, e-commerce, streaming, energy, and social. None of it was planned. But it left me with a suspicion I keep finding evidence for: most hard engineering problems turn out to be translation problems between people who do not share a vocabulary. A good part of my job is standing in that gap with a whiteboard.
These days the work splits in two directions. One is a social platform being built for the 2028 Olympics. The other is a GSA-cleared government contractor whose project list reads a little like a dare: datacenter buildouts, off-grid power systems, robotics, graphene-based materials, federal RFPs, and, inevitably, software.
Alongside that I run my own research in applied ML for energy systems: demand flexibility for AI datacenters, battery life prediction that transfers across cell chemistries, and forecasting that stays calibrated when the distribution shifts underneath it. With a friend at UCLA I also work on AI safety and ethics. Both are being written toward peer-reviewed publication. Neither is there yet, and getting there is proving to be a slower and more humbling education than the research itself.
When I am not doing that, I am usually building RC cars or outside doing something. Ideally both at once.
