Email
My Project
Status
What I've picked up along the way about product development, AI, and building things that actually work.
The 2026 panic about the dying junior developer role assumes one path into software. There has always been a third — and AI tools amplify it.
Most landing pages don't position. They apologize. People who know what they are don't hand out this many credentials.
In a year of agentic everything, the real decision is what stays in your hands.
Good design doesn't disappear. It knows what to show, and what to hold back.
Most websites in 2026 look like PDFs. They could be places instead.
With a client I can say no. With myself, I always say yes — and that's where the trouble starts.
The demo always works. The problem is what comes next, when someone drops in their own input and the magic quietly falls apart in real time.
Everyone writes about what AI takes. Almost nobody mentions what it returned.
Nobody reads your CV. Your website is the real resume — and most developers ship it once and never touch it again, like a school assignment.
Everyone wants AI in their product. Few ask whether the product works without it.
At 2 AM, the world shuts up. That's when I can finally think.
Products aren't shaped by possibilities. They're shaped by limitations.
Agentic engineering isn't about using AI to code. It's about having your marketer, designer, analyst, and developer in a single system that knows everything.
You decide everything — but someone else always picks up the tab.
A beautiful interface isn't design. Design is the decision behind the interface.
Polishing isn't high standards — it's avoiding the verdict.
When you love the work, stopping is the hardest part. Rest isn't what happens when you're tired — it's a skill you have to practice like any other.
Every best practice was born from a specific problem. The question is whether that problem is yours.
You ship it, you refresh the dashboard, and nothing happens. The world doesn't owe you attention.
As a solo founder, you never leave work. Not because you can't stop — because there's nothing to separate from.
Simple solutions look trivial. Complex ones look smart. That's the trap.
AI executes everything. But it can't decide what should exist.
Everyone's chasing the perfect prompt. The real question is what the model sees — and what it doesn't.
Design isn't how it looks. Design is how it works — and why it's worth it.
When strategy, design, and code live in the same head — less gets lost, and more gets shipped.
The hardest decision isn't what to build. It's what not to.
Speed is a multiplier. What it multiplies is the quality of your thinking.
