I direct the Applied Science efforts at Actual. While this role is wide in breadth my two primary areas of focus are in (1) delivering data and modeling software to our brilliant customers that empowers them to invest in massive physical transformations within their organizations and (2) scaling our platform to meet the needs of an incredibly wide set of use cases. My day to day includes everything from writing code to leading a small group of engineers and scientists to training our customers in our software. At one point or another I have worn the hats of every role you would expect to find at an enterprise SaaS startup other than outbound sales.

I led the development of the airspace digital twin product at Acubed, taking it from a research idea that was used to publish academic papers to an enterprise product used by 100s of scientists, engineers, and regulators around the world. The product continues to grow in adoption, and is used to evaluate and shape future aviation technology and policy. Along the way, I published a number of academic papers (some of which were privileged recipients of best paper awards), led a dozen engineers working across the globe to productize this complex software, and scaled both the technology and adoption of this innovative product.

You read more about the airspace digital twin and my time at Airbus from the Acubed blogs - more about me and more about the airspace digital twin.

I built out the infrastructure and deep learning models for drug discovery. I was the founding engineer for about 9 month, and set the engineering processes and structure in place along with the built out of the MVP for the company. The year was 2017, right before Attention is All You Need was published. I ended up using GANs for this work, with moderate success. My work helped get incubated and raise a Series A.

I spent three years at the Stanford Intelligent Systems Laboratory working with Mykel Kochenderfer on deep reinforcement learning, probabilistic decision making, and multi-agent systems. I am technically on leave from my Ph.D. program - perhaps I’ll return someday.

I look back on my time at Stanford fondly - amongst the highlights were the collaborations with lab-mates and mentorship of undergraduate students; best paper awards; starting the first multi-purpose open-source framework for partially observable Markov Decision Processes in Julia; and of course ultimate frisbee.

I miss the creative melting pot of bright minds and the constant push to further human knowledge.

I studied physics at Cal. My early physics foundations have influenced how approach problems deeply. I strongly believe in first principles thinking, and lean heavily on data to support my hypotheses and conclusions.

As an undergrad, I worked on the NEXT Experiment at the Lawrence Berkeley National Laboratory. NEXT is an incredibly ambitious project aiming to find a hypothetical (and elusive to this day) nuclear interaction called neutrino-less double beta decay. This discovery would imply that our fundamental understanding of physics needs reworking. Working on NEXT was my first experience with real-wold problem solving. I worked on nearly every part of the project - from setting u hardware calibration data pipelines, to implementing physics simulation frameworks, to soldering hardware together. I made many mistkaes, and learned a lot.