I'm a Ph.D. in Physics interested in Artificial Intelligence, building communities, compassionate leadership, and public policy. As Director of AI/ML Engineering at Dexcom, I lead the team responsible for all consumer-facing AI/ML surfaces, from research and design through production and monitoring. I spend a lot of time thinking about how Large Language Models and generative AI systems are impacting the way humans work and live, and my primary research interests include AI alignment and evaluation, large metabolic models for biosensing, and the integration of wearable data with AI systems.
I received my Ph.D. in physics at University of California, San Diego, where my dissertation research in experimental particle physics focused on new physics searches at the Large Hadron Collider, and received my undergraduate B.A. in Physics and Mathematics from University of California, Berkeley, where I worked on machine-learning optimization of physics searches at Lawrence Berkeley National Laboratory.
I'm also a big believer in staying socially engaged and politically active. In the past I have advocated on behalf of fellow graduate students, argued for increased investments in higher education and research, and implemented programs for historically underserved communities. I continue to stay close to local, state, and national advocacy initiatives through alumni groups and conversations with fellow policy junkies, and am always eager to chat about how we can work together to drive meaningful, equitable reform.
Director, AI/ML Engineering, Since March 2026
Sr. Manager, Data Products & AI, Sept. 2023 - March 2026
I lead AI/ML Engineering at Dexcom, where our team owns Dexcom's consumer AI products end-to-end: research, design, launch, and production serving at scale. Recent launches include the first Generative AI platform in Glucose Biosensing and a multimodal AI agent for meal logging serving hundreds of thousands of users. Much of our work sits at the intersection of LLMs and regulated medical software, so we spend a lot of time on evaluation, alignment, and FDA-reviewed safety strategy for production AI agents.
Sr. Manager, Experience Individualization, Dec. 2020 - Aug. 2023
Staff Data Scientist, May 2018 - Nov. 2020
I joined Dexcom as one of the data team's first data scientists, and later built up a new data science function as the leader of our Experience Individualization (XI) team. XI used commercial, medical, and app data to understand how patients actually use Dexcom devices, and turned that into personalized experiences, better in-app support, and A/B experiments across marketing and customer experience.
Doctoral Student Researcher, 2014 - 2018
My dissertation focused on new physics searches with the Compact Muon Solenoid (CMS) detector at the Large Hadron Collider, advised by Professors Frank Wuerthwein and Avi Yagil. I developed new analysis techniques and observables for CMS, and built C++ and Python frameworks for processing petabytes of collision data. It remains one of the most intellectually stimulating stretches of my career, and the place I learned how to work inside both small teams and huge collaborations.
President, 2017 - 2018
VP of External Affairs, 2016 - 2017
Legislative Liaison, 2015 - 2016
Graduate school would have been a very different experience without the community I found in the GSA. Through Lobby Corps, as Legislative Liaison, and later as VP External, I worked on our advocacy priorities and represented graduate students at the state and national level. As President, I ran the organization and its initiatives on behalf of 7,000 UCSD graduate students.
Research Assistant, 2012 - 2013
Berkeley is where I first got into particle physics, programming, and machine learning. As an undergraduate I worked with Professor Yury Kolomensky and the BaBar collaboration at LBNL, searching for new physics in invisible dark photon decays. That work is what pushed me toward a physics PhD.
A few other things I'm working on
An AI campaign assistant for D&D and other tabletop RPGs. Stop taking notes and stay in the story. Loremaster listens to your session, generates a cinematic recap, and tracks your world from one game to the next.
An automated data broker opt-out system. Uses an agent to manage privacy deletion requests for you, process their status, drive them to completion, and periodically re-poll brokers to keep your data removed.