AI systems that work without a large team.

I learned what happens when you depend too much on a single platform: 50 affiliate domains, 20 employees, expansion into three countries. Google changed the rules, Amazon shut down the partner program — and suddenly the business model was gone.

After that I made the classic mistake with Zipdo: thinking too big, burning too much, validating too little. Gave up the office, took a year off.

Both taught me something that shapes all my work today: More people don't solve the problem. Better systems do.

Today I build two AI products with two small teams — Rawshot.ai for fashion e-commerce photography and Careertrainer.ai for leadership training. Plus two to three consulting projects. No office, no VC, no agency. Instead: a custom CRM, a website that updates itself, content pipelines that run without manual effort.

What drives me is a simple belief: AI doesn't replace quality — AI enables new quality. A shop that could never afford a photo shoot now has professional product images. A leader who attended one seminar a year now trains every week. A three-person team manages a hundred landing pages. These aren't efficiency gains. They're things that were simply impossible before.