The system I wish had existed.

I wrote my first line of code in 1980 on a teletype connected to a county mainframe. I've been coding for myself ever since, using around 20 different languages over the years. Mostly, because I had a problem to solve and needed a tool that I couldn't find.

During the day, I'm a technology executive overseeing IT operations, infrastructure, security, and service delivery in regulated and enterprise environments. At night, I build privacy-first tools and frameworks through Ventura Nomadica. My focus is on turning ideas into durable, usable products that don't fall apart when the original builder walks away.

Trail came out of frustration.

While building TrekCrumbs, a privacy-first travel planning app, I didn't want to learn a new language. I leveraged AI to write the code while I acted as architect, product owner, tester, dishwasher, and cook. The process was chaos — good chaos, productive chaos, but chaos nonetheless. Decisions were stored in chat threads. Context often vanished between sessions. AI produced great results, but had no memory of why we made certain choices. When something broke or needed revisiting, reconstructing the reasoning was more difficult than it should have been.

Trail is the system I wish had existed during that build.

It's an artifact-driven framework for executing work with humans, AI, or both, without letting the conversation become the source of truth. Define what you're building with enough precision so any capable executor can pick it up from the files alone, produce the output, and leave a clear trail of why and how.

The website you're viewing was created with Trail. That's not a side note; it's the main point.