Thirty years in the trenches of networks, systems, and software.

I started in the Unix and Linux world in the mid-to-late '90s as a hacker and reverse engineer. Before that, honestly, it started with CB radios — the kind of kid who wanted to know how the signal actually got from one antenna to another, and wouldn't stop asking until someone showed him.

That curiosity paid off in an unglamorous but important way. After moving to Indianapolis, I took a job at a surplus dealer, which turned out to be the kind of place where opportunities walked in the door if you were paying attention. One of them was a spectrum analyzer I was selling on eBay. The person who bought it got me into the ISP industry, and I've spent essentially the rest of my adult life there — rolling out solutions, delivering internet to rural areas, and learning how hard it actually is to run infrastructure that people depend on every day.

In 2008, my consulting company merged with Freeside Internet Services, the long-running open source ISP billing platform. I spent the next eight years as Director of Support, working directly with ISPs on real-world billing, provisioning, and operational challenges. That work shaped how I think about software to this day: billing systems are unforgiving, provisioning has to work at three in the morning, and ISP operators have no patience for abstractions that don't hold up in production.

After leaving Freeside, I wrote a network management system from scratch and took it with me to a wireless ISP, where I served as CIO. I kept developing the NMS platform, and when it was eventually sold, I moved to the acquiring company as Chief Product Officer. For several years I ran product and engineering at a billing platform serving ISPs nationwide — three development teams, a product team, multiple production microservices, and the dual role of CPO and primary Product Owner across all of engineering. I recently stepped down from that role to build Davis Analytica full-time.

AI-assisted development is the first shift in how we write software that has genuinely surprised me in a long time. It's also the first shift where I've watched a lot of teams adopt the tools and end up worse off than they started — drowning in unmaintainable code, quality gates quietly turning yellow, architectural decisions being made by autocomplete.

It doesn't have to go that way. I've spent the last several years building my own practice around AI-native development — context files, skill libraries, quality gates, review discipline, agent patterns — and the code my team and I ship today passes A-rated SonarQube standards consistently. It's faster and better, but only because of the system around the tools, not the tools themselves.

Davis Analytica is where I help other engineering organizations build that system. I stepped down from my CPO role to do this work full-time because I believe the window to get AI-assisted development right — before bad habits calcify into unmaintainable codebases across the industry — is narrower than most leaders realize. The consulting work is grounded in products I've shipped, engineering organizations I've run, and quality standards I can prove.