IA & NumériqueLes achats après le bruit.

There's a version of "AI in procurement" being sold in conference keynotes that doesn't survive contact with a real category manager's day. It promises end-to-end autonomy, zero-touch sourcing, agents that negotiate while you sleep. Then the user opens their tool and the agent has booked the wrong supplier, ignored the framework agreement, and burned through the model budget for the quarter in three weeks.

That's the gap this series exists to close.

The Procurementor's AI & Digital series is built for procurement leaders who have to make AI and digital tools actually work — under a budget, against a roadmap, with a team that has more pressing fires than another pilot. It's not anti-AI. It's anti-fluff. There is enormous leverage in what's now possible: cost intelligence at speed no human can match, contract review at scale, conversational front-doors for stakeholders who never wanted to learn another tool. The leverage is real; the playbook is mostly missing.

I've spent the last decade deploying procurement technology in environments where the consequences are immediate. At Amazon, that meant rolling out Coupa across 800+ buildings with energy contracts that flexed monthly. At Cellnex, it meant standing up a French procurement function from a near-blank page while picking the right tools for a €200M spend. As an independent, it's meant Ariba Guided Buying deployments for CAC40 finance teams, IFS ERP build-outs for industrials, and now — increasingly — helping procurement leaders work out what to do with agents and large language models that can read a contract in 30 seconds but can't tell you whether they should.

The four sub-series each tackle a different slice:

Managing AI Cost. The category most procurement leaders haven't built yet. AI bills behave like cloud bills did in 2015 — they look small until they don't, and by the time finance asks you've lost the visibility window. This sub-series covers token economics, model selection trade-offs, the FinOps playbook for AI specifically, and how to build the controls before you need them.

AI & S2P. Where AI lives inside the source-to-pay process — sourcing intake, RFP authoring, contract analysis, supplier risk monitoring, payables matching. This is where the cluster of high-leverage, low-risk wins sit, and it's also where it's easiest to get sold something that doesn't work. The HR-process-for-AI-agents framing is the lens.

Digital Procurement. The non-AI side of digital — pilots that actually deploy, hypercare phases that don't collapse, productivity killers that no software can fix, and the operational habits (office hours, intake, governance) that determine whether your digital stack creates value or just exists.

Experimentations. A deliberately empty shelf for now. As I run my own experiments — building a contract analyser, testing an agent, breaking something in production — the field notes go here. Watch this space.

If you're building, buying, or rebuilding a digital procurement stack, this is the body of work to read first. Start with whichever sub-series matches your current quarter's biggest fire.