Evan Bunnage

I live in Seattle and recently worked as a software engineer at Research Triangle Institute in the Center for Data Science & AI.

Before that I worked at Absci, a biotech organization that uses AI to design new antibodies. I also spent some time as a consultant helping biotech startups design better proteins with machine learning. In college I studied biochemistry.

I'm interested in building software that makes machine learning in the sciences more accessible, performant, and reproducible for scientists. I also love experimenting with ways that natural language can power radically new user experiences. I think there is a lot to be excited about where these two overlap.

When not in front of a computer I like to fly small planes, ski mountaineer, and work on small construction projects with friends and family.

GitHub

LinkedIn [email protected]

Things I’ve made

my amateur carpentry
stash

A Zig library for creating binary data formats with zero-copy reads.
Elucidating disulfide bonds in biologics through semi-specific proteolysis with thermolysin and LC‑MS/MS analysis · CASSS Mass Spec 2018 poster

Verifying disulfide bonds in antibodies usually requires chemical denaturants and an overnight digest. In work done at Absci, we showed that a heat-stable protease, thermolysin, fully digests trastuzumab above its melting temperature in under 2 hours without denaturants for disulfide mapping by LC‑MS/MS. See recent work on the same problem.

Works of Questionable Rigor (i.e. select vibecoded projects)

Colorwork chart designer for knitters with WebMCP support
Command-Tab between windows on macOS