If You Can Build at Team-Scale Alone, the Bottleneck Isn't the Work Anymore

I'm not someone who posts a lot online. I've always been more comfortable doing the work than talking about it. If you'd told me five years ago I'd be publishing a series about what I'm building and how, I'd have found that unlikely.

So why start now?

Because I believe something fundamental is shifting in how software gets built, and that shift changes who can build what — and how those people need to show up.

The rise of AI agents that can interact with real engineering toolchains — Jira, GitHub, AWS, the lot — means that a single person with the right experience and judgement can produce work that used to require a team. The barriers to building serious software are dropping. Not because the work is getting easier, but because the leverage available to one person is getting dramatically better.

I've been living that. I run Estyn Software, an enterprise architecture and digital transformation consultancy based in southern California. Lately I've been helping people out with application and integration ideas they have, and building a common component that keeps coming up across those projects. It felt like the right opportunity to start journalling about what it's actually like to build something like that with AI — the process, the toolchain, what works and what doesn't.

That's a way of working that didn't exist two years ago.

Here's the thing about this world, though: if you can build at team-scale as an individual, the bottleneck shifts. It's no longer about whether you can do the work. It's about whether anyone knows you exist.

That's uncomfortable for people like me. I'd rather let the work speak for itself. But in a market where more people can produce quality work independently, the ones who document what they're doing — who make their thinking visible, who show how they approach problems — are the ones who get found. Not because self-promotion is suddenly virtuous, but because it's becoming necessary. The independent consultant's portfolio isn't a CV anymore. It's a body of visible work and visible thinking.

So that's what I'm doing. I've started a blog — not a tutorial series, not content marketing — documenting the journey of building Estyn honestly. The architecture decisions, the ways of working, the mistakes. Starting with Porth as the first real case study: why I chose to build rather than buy, how the AI-agent workflow actually operates, and what the limits are.

If you're in a similar position — someone who's good at the work but not naturally inclined to talk about it — maybe this will resonate. Or at the very least, maybe seeing someone else wrestle with the same reluctance will make it feel a bit more approachable.