Hi, Goutham here. I’m a software engineer specializing in the design, analysis, and construction of data-intensive software systems for mathematical modeling, simulation, and optimization.

What does this mean? How input data is prepared & served, how math is implemented on top of the data, and how to rigorously measure and adjust usefulness are all fair game for me. Stylistically, I love working backwards from organizational goals to figure out what needs to be done and how best to enable my colleagues in end-to-end execution.

I work with both startups and with smaller teams at large companies through my consultancy, DataTailor, for larger time-bounded projects, and individually for longer-term, combined strategic and technology advisory services.

I’ve also pursued this work as an AI/ML engineer within Big Tech. Most recently, at Meta’s Reality Labs, where I led multiple teams against the biggest bottleneck for our ML-related research into AR input & interaction - better creation and utilization of curated data in service of fine-tuneing foundation vision language models, adding supplementary input modalities, and increasing alignment with human feedback. And before that, at Apple, where I developed on-device ML models, including leading the transition of the single tap Smart Reply feature on the watch to a (much better) Transformer-based model.

I’m based out of Seattle, WA. Outside of work, you can usually find me jogging, singing in a choir, practicing my South Indian cooking skills on my patient wife, or trying to formulate my own understanding of various topics and bits of research in mathematics, statistics, computer science, & (lately) physics. Once upon a time, I worked on deanonymizing early Bitcoin transactors, and I graduated from Harvard College in 2007 with a degree in Mathematics, where I researched mathematical properties of population evolution in networks.