He gained 15 years of career over a coffee
I show him a roundtable of embodied AI agents. A Lean-Agile expert. A strategic buyer. A financial controller. Debating his case in front of him, in real time.
His exact words: "I'm going to reach a level of discussion and expertise I could never have had."
What happened in the minutes that followed is worth more than the demonstration. He pulled out a business case that had been sitting on his desk for three weeks, a funding candidate he could not challenge on his own.
The strategic buyer immediately questioned two cost assumptions. The financial controller pointed out an inconsistency in the 18-month ROI. The Lean-Agile expert reframed the problem as a value stream. He, in the middle, was no longer producing the analysis: he was arbitrating.
In fifteen minutes, he obtained what would have taken him two steering committees and three rounds of back-and-forth. Without being a finance expert, without mastering strategic purchasing, without fifteen years of Lean-Agile behind him.
The power of convening
This is where his sentence takes on its meaning, and it says something other than the usual story. He is not climbing the learning curve of three professions at once — that would be impossible and he does not claim it. He is accessing the conversation that would have taken place between three experts around a real table, if he had held the power to convene them on a Tuesday morning for his own case.
That power of convening is the real break. Not content production, not writing speed: access to a level of arbitration that until now was reserved for those with the network, the seniority or the budget to mobilize it.
People keep saying AI will replace experts. In the field, I see the opposite happen: an operations lead walking out of a coffee with permanent access to arbitrations he could never have afforded, even at the end of his career, and a case he can finally take to the committee without a crutch.
4 hours a week, and that was not the real problem
The same profile, on another engagement, works across fifteen different tools. The ticketing system, project time bookings, forecast spreadsheets. Information arrives late, scattered. One synchronization error and millions of euros of external commitments drift.
Before, he reconstituted the summaries by hand. Nobody else knew how. He played data detective while the strategic drifts slipped under his radar.
He put the real cost into words himself: "I see the drifts three weeks too late, every time, when they show up in the figures." He never had the time to look up and raise the alarm.
An operations lead spending his days reconstituting data is a strategic brain turned into a human ETL.
The real scandal is not the 4 weekly hours. It is the decision that was not taken, the arbitration that slipped under the radar, the budget drift he could have stopped at €300K and that ended at €1.8M. His expertise was costing the company because it was deployed in the wrong place.
I did not deliver him a solution
I taught him, in one hour and a coffee, to structure it himself. His consolidated summary now arrives in 3 minutes, and he can make it evolve on his own when his scope changes.
The distinction is not cosmetic. A delivered solution creates a dependency: at the first scope change, someone has to be called back. A transmitted capability survives the change, and it is the only deliverable that keeps producing value after I leave.
That is what an AI transformation is. Not automation, and not generic training: support that starts from that person's real irritants.
The double channel, collective and individual
This level of precision is not obtained in plenary, and it is the second thing these engagements taught me.
You launch your transformation. The collective is aligned, the rituals are in place, the slides are approved. Three months later, the figures barely move. What was missing was not in the room: it was in what nobody dared say in plenary.
On a recent engagement, I split the support in two. Collective on one side, individual running in parallel on the other. The individual sessions served one purpose only: collecting the field's fears unfiltered, anonymizing them, then sending them back into the collective reworked as actionable subjects.
That double channel changes the nature of the transformation. The collective pushes communication. The individual surfaces what blocks it. One without the other, you are running theatre. The two together, you transform.
This is what tailored support means — not a premium option on a catalogue, but the condition for anything at all to reach the ground.
Listening before proposing
One last case, because it holds for the whole approach. "Our needs are already covered, come back and see us in 2027." That was the first sentence I heard on arriving at a pre-sales meeting — the kind that closes a discussion before it starts.
Three quarters of an hour later, the same person was asking whether we could meet again in a month.
What happened in between was not a technique: I started listening. His real constraints, his existing ecosystem, the one I would have to fit into without breaking anything. What they actually wanted, rather than what I felt like selling them. The shift happened at the precise moment I stopped pushing a solution: they started pulling.
Staying tool-agnostic is not a commercial posture. It is what turns one more vendor into someone you want to keep talking to. The "come back in 2027" was not a refusal, it was a test to see whether I would insist or whether I would understand.
These four situations tell the same story. The value of AI in a company does not sit in what it produces, but in the conversations it makes possible and the decisions it allows to be taken in time — the observation I develop in my field report on augmented agile transformation.