I'm a computer science PhD candidate at UMass Amherst working with Don Towsley at the intersection of machine learning and network science. I've interned at Google as a software engineer with both research and product teams and at Technicolor Labs as a researcher.
Before school, I was an Associate Scientist at Archimedes, where I designed and implemented computational disease models; before that, I earned a Bachelors in Physics from Oberlin College with summer jobs hacking Portage, working in an observatory, and shoveling, um, dirt.
We recently developed a neat neural net for graph data that was accepted at NIPS 2016 and we've released our source. We're also working on some data-oriented privacy-preservation stuff, although that's still under heavy development.
I keep an informal blog that is mostly just rants about how hard it is to do anything. My 2016 Google internship project was released as an open-source TensorFlow contribution. Here's a lisp interpreter - it interprets lisp. Here's a thing that renders greek symbols (or whatever) in emacs as you type. Here's the source for this site. It's really just a Jekyll clone - you should probably use that instead. Here are my github and linkedin profiles, and here's my resume. You can also find me on twitter at @jatwood.
I'll be graduating from UMass in Spring 2017 and returning to Google Cambridge full-time. You can reach me if you aren't a robot and know what james.c.atwoodREMOVETHESECAPS@gmail.com means.