WRITING & NOTES

Think it through.
Then build.

Perspectives on software, AI, and the responsibility that comes with making technical decisions.

FIELD NOTES / 2026

The latest writing.

Longer pieces on delivery, AI-assisted engineering, and the judgement that has to sit behind both.

3 MARCH 2026 · 1,013 WORDS

Execution as Architecture: Engineering Closer to Reality

Nobody sets out to slow progress down. Alignment meetings, architecture reviews, risk controls and sprint planning each make sense on their own, but together they stretch momentum thin. Faster to Value is about creating space where execution happens closer to the moment an idea appears.

Read the piece
27 FEBRUARY 2026 · 2,060 WORDS

When the Robot Isn’t the Bottleneck… You Are

I can manage one or two deep AI workflows at a time if I start early. Beyond that, my brain pushes back — not because the AI is underperforming, but because the accountability never switches off. Prompting feels like an early interface, not a permanent one.

Read the piece
23 FEBRUARY 2026 · 792 WORDS

The Road to the Dark Software Factory, and Where I Actually Sit on That Highway

Somewhere between spicy autocomplete and the dark factory sits an honest question: where are we actually right now? For me it is rails everywhere — agent contracts, policy gates, architecture validation — and a human still carrying the decision.

Read the piece

IDEAS FROM MY RECENT LINKEDIN WRITING

Shorter notes.

NOTE / 01

AI output still needs an owner.

AI can challenge your thinking and help you structure feedback. Before it becomes your advice, read it, verify it, and decide whether you stand behind it.

NOTE / 02

Better AI use starts with cost awareness.

More tokens and more agents are not automatically better engineering. Use the tools where they create value, with attention to budget, context, and the work that follows.

NOTE / 03

Get closer to real value.

Build early, release responsibly, and learn from actual use. Systems thinking and clarity of direction matter as much as the speed of producing code.

Follow my writing on LinkedIn ↗

THE ARCHIVE / 2019

Earlier thinking.

Original articles, kept at their existing addresses. Some opinions and project details belong to the time they were written.