BSE, Biomedical Engineering, Duke University, 2012
Biology-driven products for an exciting future.
Neuralink is developing high-bandwidth brain machine interfaces, initially for critical unmet medical needs including paralysis from spinal cord injury, and eventually for merging minds and machines.
Having grown up programming, where people invent whole new languages when they get tired of typing too many semicolons, the sheer inefficiency of pressing a physical button on the plate reader every three hours for two days to gather the data I needed struck me pretty quickly. Automation had been around for biology for a long time, but it was prohibitively expensive and difficult to use. For that matter, all lab equipment was prohibitively expensive and difficult to use.
Borrowing the metaphor from what cloud computing had done to software, the idea of creating a robotic cloud laboratory for biology seemed obvious. I founded Transcriptic right after I graduated from college and over my five years as CEO raised over $30M from Google Ventures, IA Ventures, Data Collective, Silicon Valley Bank, and others; and grew the business to millions of dollars in annual revenue.
I stepped down as CEO in 2017 to co-found Neuralink but remained involved with the company through early 2019 when it merged with 3scan and rebranded as Strateos.
The college admissions process can be modelled as a sequential move game with simultaneous move subgames, from both the student and schools' perspectives. MyFit originally started as a Facebook app I wrote to gather data and build models my friends and I used to reason about our own college admissions chances, which eventually turned into a venture backed company.
After the intial focus on modeling admissions chances, we realized that if you could predict the statistical lift in lifetime earnings from a degree for a matriculating student, you could underwrite education using an equity instrument rather than debt, which would have all kinds of exciting downstream implications. At the same time, colleges have challenging yield forecasting problems that demanded more sophisticated modeling than was widely used at the time. I took a year leave of absence after sophomore year of college after raising a venture round and moved to California to build the company. We were ultimately unsuccessful at revolutionizing education finance, but learned a lot and ended up selling the technology to the big company that made the software colleges used to manage their admissions process.
For as long as I can remember, my ultimate life goal has been to get a brain-machine interface and disappear into the Matrix. I was fortunate that Duke had, at the time, one of the best neural engineering labs in the world. Having worked my way into the Nicolelis lab through an independent study during my freshman year, I proceeded to spend most of my time in college there learning everything I could.
I presented a poster at the Society for Neuroscience Annual Meeting as a sophomore in 2008 and my main project for much of my time in the lab was studying how differing mappings between joystick-space and screen-space in an upper limb task for macaques led to changes in neural representations. I was only seriously derailed by silently broken timestamps in the ephys data once during my whole time in the lab.
Computer Chips in Brains Could be the Future of Medicine (TIME, November 2025) (video)
Can The Brain Merge With Artificial Intelligence? (Alex Kantrowitz at Web Summit, December 2025)
After Neuralink, Max Hodak is building something even wilder (TechCrunch, December 2025)
Ex-Neuralink Exec's New Startup Is Tackling Blindness—and Winning (Inc., October 2025)
Brain Computer Interfaces, Explained (Abundance Institute, October 2025)
Restoring sight is possible now with optogenetics (Washington Post, April 2024)