Writing
Articles
A personal view of the professional landscape we work in.
Inspecting more is not the point
Robotic inspection collects more data, and more precise data, than anyone could before. But collecting it is only half the problem. The other half is turning it into something a decision can rest on.

Your people already use AI. Under what controls?
I do not usually recommend software. This is one of the exceptions: it solves a problem I see everywhere in small and mid-sized companies, and — worse — in some very large ones too.

AI does not think like us: why humanising it is a mistake
“AI will take our jobs.” “AI is conscious.” “AGI is here.” Really? After 30 years in software engineering, the hype is getting out of hand.

What if we are not imitating it well?
Sometimes, to open a new path, changing the question matters more than perfecting the answer. What if today's AI is still a very early and very crude imitation of nature?

Not everything that glitters in AI is gold
A new model every week, a record every month, a promise of disruption every quarter. But hype is foam, and foam weighs nothing. What matters to a company sits underneath.

When the connection disappears, only control remains
Passing behind the far side of the Moon, the Artemis II crew spends 40 minutes with no link to Earth. Only what is aboard remains. Enterprise AI works the same way.

One Claude is no longer enough. Orchestrating agent teams in Claude Code
Imagine cloning your best engineer five times, each taking a different angle on the same problem, and having them argue until they converge. That is Agent Teams, and you can use it today.

AI does not replace engineering. It makes it more necessary than ever.
Plenty of companies want to add AI to their products, and very few are ready to integrate it properly. What separates a flashy demo from a sustainable product is engineering.
Let's Build Something Great
Have a project in mind? I'd love to hear about it and explore how I can help.