Antoine
Craske

Before AI agents became part of the engineering conversation, I was already working on the underlying problem: how technology organisations increase their capacity to turn intent into reliable business change.

At La Redoute, I helped transform an e-commerce organisation moving from large quarterly releases and extensive firefighting toward daily deployment and highly automated quality engineering. The transformation spanned CI/CD, trunk-based development, cloud-native architecture, APIs, event-driven systems, observability and organisational change.

Quality Engineering became a particular focus. I developed the foundations that evolved into Cerberus Testing and helped scale automated functional, integration and end-to-end validation to more than 7,500 tests, reliably supporting daily deployments across countries.

But the most important lesson was not about testing.

Every time we removed a constraint, another one became visible.

I’ve since repeated variations of this transformation across manufacturing, retail, energy, wholesale and financial services — changing engineering organisations, architectures, data flows and technology operating models to increase their ability to deliver and absorb change.

Technology problems are almost always system problems.

The interesting question is rarely “which tool?” but “what does this change in the system — what constraint does it remove, and where does the bottleneck move next?”

AI is now creating the largest increase in software production capacity I have seen. That makes the question of constraint migration more important than ever.

That lens drives my central project: redesigning engineering production for the agentic era. I call the space between AI’s production supply and an organisation’s ability to safely absorb it the Absorption Gap, and model the whole flow as the Engineering Production System.

I am the author of The Systemic CTO, The Systemic CTO in Action, the AI Gap series and Quality Engineering — alongside 300+ articles. My frameworks include MAMOS, AAA/3A, MVM, Quality at Speed, the Seven Gaps and the AI Gap 100-Point Assessment.

QE Unit has been consolidated into this site. Its core Quality Engineering writing now lives here, re-read through the lens of AI Engineering Transformation.

Elsewhere: LinkedIn · X · GitHub · Leanpub