From early web projects to engineering leadership
My father started me on my journey into technology. He wasn’t a techie, but he bought me a computer and made me promise to learn how to use it and to create something. It wasn’t until I bought my own computer that I started working toward that promise. I soon found myself contributing to open-source projects. One of my PHP3-era osCommerce add-ons, NewsDesk, was popular at the time and was later bundled into projects like CRE Loaded.
My PHP journey has given me many wonderful memories and experiences. Since then I’ve moved my way up the experience path and met many wonderful people, even a few who became good friends.
Being a developer at heart, you always spend time learning new frameworks, languages, and tools. My dad urged me to build things like a self-driving car. Obviously, I never did learn how to build a car that could drive itself, but my fate led me to building things with AI instead. The real journey ran from Perl and flat-file databases, asking questions like, “What is ?”, to leading developer teams, to where I am today: building a governance tool for AI agents.
Why Hakama exists
Hakama started from my frustration with AI agents: scope creep, drift, ignoring prompts and skills, and agents wanting to DIY new patterns. I joke that I ask AI to “paint the living room grey”, and the agent instead paints my kitchen green, feeds my dogs, and generously decides I need a rainshower head installed in my shower.
Of course, I tried prompts, then skills, then structured workflows, frameworks, and loops. AI agents still ended up telling me, “Sorry, Bro, I should have read the full file.” I finally had my lightbulb moment after one too many of those apologies. I sat down to work within the confines these agents give you and find what actually keeps them in line. Hakama was born from that need: to not only have AI do the coding, but to have it pay attention to what humans want too.
Built for the whole team
Part of my goal was that I didn’t want to build just another developer tool. I lived through the years of “another XYZ open-source project.” I focused on the whole team: a tool that a code reviewer, QA, product owner, or engineering lead could turn to, so the moment a code push lands the whole team has a real-time view into the code, documentation, and other telemetry. The benefit is that Hakama cuts down on onboarding and the endless developer and non-developer meetings. If AI agents can make developers more productive, then they should do the same for the whole team.
Why the name Hakama
I named this AI governance system Hakama because I’ve been a fan of martial arts since some of my earliest memories. I’d go to sleep early just to wake up for the late-night Kung Fu Theater movies. I also speak Japanese, and Japanese names have a way of following my work: I was inadvertently the inspiration for the naming of Zen Cart. A Japanese name felt only natural for this AI governance system, and the Hakama name is a fitting homage to the golden age of rockstar and ninja programmers.
If you want to talk about governing AI-assisted change on your team, request a pilot or reach out directly.