My journey to Arena Physica began with a double major in physics and philosophy, after which I went straight into a Ph.D. in Applied Physics and Scientific Machine Learning. I've always been fascinated by the intersection of physics, math, and computer science, and during my Ph.D. I developed machine learning models to solve forward and inverse design problems in applied physics and engineering. When I finished my Ph.D., I joined Arena Physica’s EM foundation model team.
I'm a machine learning research scientist at Arena Physica working on our EM foundation model. Most of my days are spent formulating and testing new hypotheses aimed at improving our model’s EM modeling capabilities. It's a challenging role, but I love it. Our research team's approach revolves around testing a large number of ideas and attacking problems from several different angles, learning from both the successes and the failures. In practice, my work ranges from implementing new variants of our model architecture to building more efficient training pipelines.
When I began looking at AI companies in the engineering space, I noticed that many were approaching EM problems by wrapping an LLM around a traditional solver. What stood out about Arena Physica’s mission was the willingness to challenge the dominant paradigm and build a foundation model trained directly on EM fields. The idea immediately resonated with me: in the same way that text teaches an LLM the structure of human language, fields can teach our model the structure of the electromagnetic world. Coming from a Ph.D. in physics-informed machine learning, I had seen how much impact AI can have in engineering, and I wanted to be part of that revolution from the inside.
Our foundation model has made real progress over the past few months, and I'm excited about where it's heading. In particular, our recent work on modeling the EM behavior of 3D devices, including printed circuit boards and antennas, has been a highlight. As the model's precision and generalization improve, I'm looking forward to seeing it deployed on real engineering problems.
One of my favorite memories is an odd one, since it dates back to my onsite interview in New York, before I had even joined. Over the course of the day I spoke with a lot of people at Arena Physica, including several members of the EM foundation model team. What struck me was that everyone combined real depth of knowledge with genuine curiosity: they were interested in what I had worked on, and when the conversation turned to ML models or EM design methods, they were open to the ideas I was proposing rather than just evaluating them. After I left the office I knew that Arena Physica was the right place for me.
My advice for anyone interested in joining us is to not be afraid of taking risks, and to trust your gut. Everyone at Arena Physica is working toward the same shared goal, so if you have an idea that can move things forward, speak up and your teammates will help you develop it into something that can be shipped. We're actively hiring for additional Research Scientists - you can learn more at arenaphysica.com/careers.
And in classic Arena Physica tradition of sharing a hot take, here's mine: tennis is the hardest sport to turn pro in!