I tell computers what to do. Most of the time, they listen.
As a developer, I’m particularly good at solving hard architectural problems and figuring out how to simplify complex systems. As a leader, I enjoy building organizations that let the team focus on building products and insulate them from organizational noise. I also seem to be very good at turning around broken teams—that’s deeply rewarding work, although it’s also very stressful.
I’ve had the good fortune to be part of several once-in-a-lifetime projects that had a major impact on the world.
2025 -
As of April 2025, I’m shifting to working on AI. I don’t yet know what I’ll be doing, though I’m particularly interested in alignment and capability assessment.
2006 - 2025
I’ve worked for myself for the past 20 years, doing a combination of commercial software, passion projects, and contract work. A couple of the projects that I’m able to talk about:
TextCarver is a platform for creating complex documents using code. It’s particularly useful for programmatically building multiple versions of a document, as well as generating documents from structured data. You can think of it as a CMS with an API instead of a graphical front end.
I spent a few years doing iPhone development. As well as some consulting projects, I had two apps in the app store (note that both names are now used by unrelated apps):
I spent a few years exploring neural networks for handwriting recognition. The work was fun, but ultimately didn’t go anywhere.
2001 - 2006
Valve is a profoundly unusual company that was consistently exhilirating to work at (and occasionally utterly maddening). I worked with some incredibly smart people, had more fun than should be allowed, and was able to have a real impact on the games industry.
I took over the Steam team while it was in beta. At that time it was evident that Steam had the potential to revolutionize the economics of the games industry, but it was held back by serious technical and organizational challenges. I created a new architecture for the project and was able to guide it to a successful launch.
My original project at Valve was to tackle online cheating in Counter-Strike. I created the original VAC system and spent several years in an arms race against an increasingly sophisticated ecosystem of cheating tools.
1988 - 2000
Early Microsoft (I was employee 4,136) was a very different company than it is today. We worked on small teams, moved fast, and left our mark on the industry.
I was development manager for the simulation games group, managing about 50 developers working on four products including Flight Simulator 2000 and Combat Flight Simulator.
I became the development manager for the charting team when it split off from the rest of the Excel development team. We remained part of the larger Excel organization although our code also shipped with Word and Powerpoint.
The early days of Excel were a ton of fun: when I started, Excel was still lagging Lotus 1-2-3 and the world wasn’t yet sold on the value of graphical user interfaces for serious work. I was a development lead, leading the charting team.
I went to high school at Georgetown Day School in DC and studied physics at the University of Chicago (although I spent most of my time in the astronomy department).
I don’t use social media.
Email’s the best way to reach me. I’m mattb (you already know the domain).