Welcome

About

Howdy! I'm an engineer working on brain-computer interfaces at Zyphra in downtown San Francisco. Currently, I am building foundation models for neuroscience, and my long term goal is to build computational replicas of human brains to build new modes of interactions with computers, diagnose and treat diseases, and understand how consciousness arises from matter.

I studied Mathematics and Statistics at UNC-Chapel Hill, where I began my research in computer vision for autonomous vehicles.

I then moved to Boston to study Computational Engineering at Harvard University, conducting research in the CRISP Group where I developed a passion for computational neuroscience.

I started a PhD in Statistics at Columbia University, but put my academic aspirations on hold to play the startup game. In January 2026, I moved from Manhattan to San Francisco for this next phase. Always happy to connect with new and old friends in the city.

Please reach out if you are interested in talking about my work or collaborating on something! I run a lot of weekend projects and we frequently submit to NeurIPS/ICML/ICLR main conferences and workshops. Any passionate undergrads/master's/PhD students are welcome to join our group. We have compute and storage for you, no need to bring anything but enthusiasm.

Personal

Originally, I'm from Durham, North Carolina. In my free time, I enjoy photography and regularly post my work on flickr. I've also put a few of my favorites in the gallery section of this site.

I'm very passionate about movies and photography. I've watched way too many films and occasionally post reviews on my letterboxd page. I think Denis Villeneuve is the best director out there making movies today. If I had to choose my all-time favorite directors, that's tough...maybe Sergio Leone or Bob Fosse?

RELEASES

~July 2026~

ZUNA1.1 is now out. Same arch as ZUNA1 but with variable-length context and more diverse dropout schemes!

~June 2026~

Completed during my PhD at Columbia, we released CASSM, ("Computation-Aware State-Space Model") a highly scalable Bayesian state-space framework for predicting single-cell neural activity. You can install it with

pip install cassm

The project site is available at cassm-torch.com

A preprint is available on arXiv and we will be giving a talk at the 2nd International Conference on Probabilistic Numerics in Finland...coming soon!

~February 2026~

A new foundation model from Zyphra! We released ZUNA, a 380-million parameter diffusion autoencoder for EEG super-resolution. You can install it with

pip install zuna

Please let me know if you have any issues or questions. You can also submit a pull request on the GitHub repository