Hi!

I am a computational mathematician / engineer / researcher excited to work on some intersection of software engineering, mathematical modeling, high-performance computating, and numerical optimization.

See my LinkedIn if you’re looking for a more formal presentation of my work and career. Else, read on!

Work Today

Im my current role at X, the moonshot factory (formerly known as Google X), I work on the Tapestry team.

Things are relatively secretive here, but I can say that I work as a software engineer on optimization and simulation for decarbonizing the electric grid.

In particular, I’ve been developing and productionizing algorithms for optimal power flow (deliver power cheaply), hydro-thermal coordination (coordinate thermal generation with hydropower generation), battery / storage systems (fill it up when power is cheap, let it out when power is expensive), and unit commitment (which generators should we turn on today?).

Past Lives

While at Cerebras Systems, I worked as a software engineer on machine learning compilers for the Wafer-Scale Engine (WSE), our big-honking-chip for (mostly) large language models. I led a small team of engineers responsible for integration of our automatic code generation library into the production compilation stack, enabling the targeted use of auto-generated kernels in situations where hand-written kernels were missing or inadequate. I spec’d out a small MLIR dialect for auto-gen kernels and worked with a team of external consultants to use this for our code generation pipeline.

At computational pathology startup Path AI between their Series B and C funding rounds I worked as senior algorithms scientist and research software lead for the imaging team. This included a large amount of algorithm development and implementation for inverse problems in computational imaging, working in tandem with imaging scientists and systems software engineers to prototype complex image processing techniques applied to whole-slide images in pathology.

In my first industry job ever, I was a software engineer at Google where I worked 80% time with the Hotels team on data analytics and 20% time with the operations research team on linear program solvers.

School

For my PhD, I studied with Lexing Ying at the Institute for Computational and Mathematical Engineering (ICME) at Stanford University.

My thesis research was focused on fast algorithms for scientific computing, in particular fast linear algebra based on data-sparse representations of rank-structured matrices that arise from physical problems in 2D or 3D. I have PDFs of my thesis and thesis defense slides.

After my PhD I was at the Flatiron Institute as a research fellow in computational neuroscience.

Previously, I did my undergraduate work in mathematics and electrical engineering at Tufts University, where my advisors were Scott MacLachlan and Douglas Preis.

My graduate funding was generously provided by the Department of Energy through the Computational Science Graduate Fellowship program (CSGF). Through them, I have had the opportunity to work at a number of DOE research laboratories:

• Argonne National Laboratory (summer 2010 with the PETSc team)
• Lawrence Livermore National Laboratory (summer 2012 with the computational mathematics group through the Cyber Defenders program)
• Lawrence Berkeley National Laboratory (summer 2014 with the Applied Numerical Algorithms Group)