A compliance attorney turned product builder.

I spent 15 years translating complex rules into workable processes in banking and public affairs. Lobbying disclosure, insider trading, pay-to-play restrictions, foreign client due diligence. I've filed the forms, built the tracking systems, and explained the rules to people who needed to follow them.

Now I'm building the tools I wish I'd had.

Tools that turn public disclosure data and regulatory knowledge into something professionals can actually use.

Four projects so far: two in compliance, two in political intelligence. Each one started with a problem I've lived.

The case studies on the Work page explain what I built, what I cut, and what I learned.

Domain expertise used to be nice-to-have. Now it's the bottleneck.

AI changed what one person can build. The constraint isn't engineering capacity anymore; it's understanding the problem well enough to design the right solution. That's the part that can't be shortcut.

I'm not a software engineer. A year ago, that would have stopped me. But the tools I'm building require something harder to acquire than coding skills: 15 years of knowing how compliance workflows actually break, what lobbyists actually need to win pitches, and why the existing tools don't work.

The people who understand these problems don't build software. The people who build software don't understand these problems. AI closed that gap for me.

I build to learn. Analysis paralysis is real for me, so I'd rather launch something imperfect and iterate than wait for certainty that never comes.

I constrain scope deliberately. A career assessment showed I'm in the 99th percentile for idea productivity, which means new ideas come constantly. That's a strength and a risk. Seven-day build challenges keep me finishing, not just starting.

I start with what's different about this user. My top CliftonStrength is Individualization: I naturally focus on what makes one person's situation different from another's. That's why I build tools for specific use cases, not generic platforms.

I learn fastest with a real problem in front of me. That's how I'm picking up JavaScript, Node.js, and the Claude API: by building tools that solve problems I actually understand.

I think about UX as wayfinding. Massimo Vignelli's New York subway map and Lance Wyman's DC Metro signage showed that good design helps people navigate intuitively. Clean Helvetica, purposeful color, no clutter. That's the standard I'm trying to meet.