I've been building things since I was a kid with a home computer. The specific things have changed — 6502 assembly, enterprise software, cloud infrastructure, AI platforms — but the compulsion hasn't. Something needs to exist that didn't exist before. That's always been the thread.
This site used to be a place for me to write about retro computing and side projects. That's still in here somewhere. But something has shifted, and I want to be honest about what it is and why.
What's Actually Happening
I'm 52. I'm CIO at Kwalee, a mobile gaming publisher, where I'm building an enterprise AI agent orchestration platform from scratch. It's serious, interesting work and I'm proud of it. But it's not mine.
Alongside that, I'm building Komputa — a product company, not a consultancy. The first product is ProductOne: a unified fitness analytics dashboard for serious lifters. It aggregates data from Hevy, Cronometer, Garmin, and Withings into one intelligent platform that finds the correlations between your training, nutrition, recovery, and body composition that siloed apps can never see.
It exists because I'm a bodybuilder frustrated by fragmented data. I have four apps that collectively know everything about my training, and none of them talk to each other. ProductOne fixes that — for me first, and then for everyone else in the same situation.
ProductTwo and ProductThree are already on the roadmap. One sector, deep. The moat is data and time.
Why AI-First Actually Means Something Here
I'm building Komputa using an AI-first methodology — not as a buzzword, but as a genuine operating model. Claude, Codex, structured agentic development pipelines. As a solo technical founder working alongside a full-time CIO role, I'm moving at a pace that genuinely shouldn't be possible by conventional means.
The AI-first approach is itself part of what I'm building in public. Every architectural decision, every pipeline choice, every place it works brilliantly and every place it falls over — I'll write about it here. Because the honest account of how this actually works is more useful than any polished case study.
Why Public
Building in public isn't a marketing strategy. It's an accountability structure and a distribution channel that doesn't require cold outreach.
The people I want to find Komputa are people who build things, who lift, who care about data, who are curious about what AI-first software development actually looks like in practice. Those people don't respond to ads. They respond to someone showing their working.
So that's what this is. The working.
What's Coming
ProductOne is in early access now. If you're a serious lifter using Hevy, Garmin, Cronometer, or Withings — or any combination — get on the list. You'll be the first to know when it launches and your feedback will shape what it becomes.
Everything else gets documented here. The builds. The decisions. The numbers, where I can share them. The methodology. The failures alongside the wins.
Ten years. One sector. Always. Building. Something.
— John
