Do I need technical skills or a data-science background to use this book?
No. The book is explicit that the AI-augmented COO needs a new working posture, not technical skill — the managerial ability to brief a task, judge the output, and verify it. If you can delegate work to a capable analyst and evaluate what comes back, you already have the core skill. There is no code to write and none to read.
Is this a book about ChatGPT?
No — it is about using generative AI in operations, whichever capable model you choose. One chapter covers model selection candidly and without vendor allegiance, including where a different model, or a self-hosted open-weights model running on your own hardware, is the better choice. The methods apply regardless of which model you standardise on.
My factory runs on machines, not text. What can AI actually change?
That question is the heart of the book. AI does not change the physical layer — the machine still cuts at the speed it cuts. It changes the cognitive layer around the machine: the analysis, the reporting, the problem-solving documentation, the decision support. The variance review that took three days, the supplier risk brief that took a week, the 8D that sat unwritten for a month — that is where the gain is, and it is large.
Does it cover the EU AI Act, GDPR, CSRD, and NIS2?
Yes, practically rather than academically. A dedicated chapter translates the governance load — the EU AI Act’s risk tiers, GDPR, CSRD disclosure, and NIS2 — into what an operations leader actually has to do, and a full domain chapter covers CSRD/ESG sustainability reporting in its current, post-Omnibus form. The regulatory material is written at a durable structural level, with the clear caveat that specifics move and should be checked against current national law.
Is this only relevant to European manufacturers?
The governance chapters are strongest for firms inside or supplying into the EU, because that is where the regulatory load is heaviest. But the operating model, the domain playbooks, the prompt library, and the hard boundaries are jurisdiction-neutral and apply to any mid-sized manufacturer. If you sell to EU customers, the supply-chain chapters are directly relevant even if you are based elsewhere.
What size of company is this written for?
Manufacturers of roughly 100 to 2,000 employees — large enough to have real operational complexity and a board to report to, small enough to have no AI team or data-science department. If that is your situation, the book is aimed squarely at you.
Will AI replace my operations people?
The book’s consistent position is that it will not, and should not. Generative AI augments operational expertise and raises its value, because expertise is exactly what stands between you and a confident, plausible, wrong answer. A full chapter addresses the human side of adoption — reskilling, the deskilling risk, and managing the workforce transition honestly — and the hard boundaries chapter is explicit that decisions about people’s livelihoods are never the machine’s to make.
What are the “hard boundaries”?
Seven things AI must never do in a factory — autonomous safety decisions, autonomous customer commitments, autonomous employee evaluation, and others. The book argues that clearly drawn boundaries do not slow AI adoption; they accelerate it, because they make it safe to move fast everywhere else.
What exactly do I get — is it just prose, or tools I can use?
Both. Alongside the chapters you get a working toolkit: a library of production-grade prompts across ten operational domains, a five-level maturity diagnostic, a data-governance reference, a governance accountability matrix, an operations strategy matrix, and a total-cost-of-ownership model you can run against your own numbers.
What’s the difference between the Amazon and Leanpub editions?
The content is the same book. The Leanpub edition is updated as the field and the regulation move, and buyers receive updates automatically — useful given how fast both AI and the EU rules are changing. Choose Leanpub if you want the living, updated version; choose Amazon if you prefer the Kindle ecosystem.
Are the examples real?
They are composites — anonymised and recombined from genuine operational situations into a single fictional manufacturer, so that real suppliers, customers, and figures are never exposed. The patterns are real; the identifying details are deliberately not.