30+ projects. Zero paying users.
That's my track record. I have a Projects folder on my computer with over thirty started-and-stopped builds. Crucible. ValidationOS. PathFinder. StratIQ. A football academy app. A youth sports directory. A monster maker for my kids. I built every single one with AI tools. Most of them work. None of them have paying users.
I'm Frank Sellhausen. I'm documenting what it actually takes to close the gap between “it works on my laptop” and “someone will pay for it” — in public, with real numbers and real failures.
I bring three lenses most builders don't have.
The drill sergeant lens.I spent 16 years in the Army, including six in a drill sergeant unit and a year as an Observer/Controller/Trainer at a Combat Training Center — where the Army developed the After Action Review. I trained units in running AARs and watched the process surface truths that rank alone never could. What was supposed to happen, what actually happened, what do we do differently next time. No ego in the room. That framework shows up in everything I build and everything I write.
The examiner lens.I left the Army to examine banks for the FDIC. I ran risk and compliance reviews — which means I spent years seeing every system through a regulator's eyes: what could go wrong, who could get hurt, where policy meets reality. When I look at an AI product, I see the access controls, the data exposure, the trust someone is placing in you when they hand you their information. Almost nobody building AI products has sat on the regulator's side of the table. I have.
The product lens.I was a Senior AI Product Manager at Dell Technologies, shipping agentic AI systems, MCP server implementations, GenAI tooling, and an account intelligence platform. Then I got laid off. I'm not going to dress that up. It happened to thousands of us, and if you're reading this, it may have happened to you too. What I learned is that knowing how to build inside a big company doesn't mean you know how to build for yourself. That's a different skill set, and it's the one I'm learning now.
What you get from following this. I'm building Long Game Athletics using the same process documented on this site. Every framework, every lesson, every mistake gets shared in real time. No guru energy, no “10x” language. Just an honest account of what works and what doesn't when you try to turn a prototype into a product.
I have four sons, an MBA from Texas A&M, and the AIGP certification in flight. I've been a hobbyist coder for years. AI tools finally let me ship. Now I'm learning the part that comes after shipping.
Real builds, real mistakes, real numbers. Written for experienced operators building their second act.