I Was on GitLab’s “Meet the Tanukis” Podcast
GitLab has an internal video series called Meet the Tanukis, where engineers and leaders share their stories. I was recently a guest, and the conversation was genuinely fun.
I talked about how I got into coding, which traces back to hot-swapping game cartridges as a kid, trying to figure out why that sometimes worked and sometimes didn’t. That curiosity turned into a computer science degree, and eventually I became the second developer at a startup and worked my way up to CTO.
At GitLab, I led three teams: Security, Applied ML, and Growth. The Applied ML work I’m most excited about right now is using machine learning to recommend the right reviewers for code changes. It sounds simple, but getting it right has a real impact on how fast code ships and how well knowledge spreads across a codebase. The Anti-abuse team is doing similarly interesting work, keeping spam and CI abuse in check at GitLab’s scale.
The Growth team came up a lot in the conversation. Growth is often misunderstood — people assume it’s all about marketing analytics and dark patterns. It’s not. We run experiments, and one thing I said that I believe pretty strongly: a failed experiment isn’t a bad thing. It’s data. If you’re not failing some percentage of the time, you’re not running real experiments.
We also got into the distinction between the Secure and Defend stages, and how I think about balancing risk across all three teams. That part of the conversation I’d actually recommend listening to if you’re an engineering leader who manages teams with different risk profiles.
Leave a Reply