A European Pharma had designed a new cluster operating model to bring more alignment across countries and functions, while interest in AI was growing at the same time. The structure was in place, roles were mapped, governance was outlined. What had not been tested was how any of it would behave once real work, real pressure, and cross-country complexity came into play. In a system where decision-making was still unclear, ownership was fragmented, and work did not move cleanly across teams, AI quickly added volume: more content, more inputs, more opinions, without moving things forward.
We stepped in as an adoption partner to make the model work in practice. We stress-tested it through real scenarios, surfaced where it would break, and rebuilt the fundamentals: who decides what, how work moves across pods, squads, and cockpit, and where accountability sits. In parallel, AI was used to accelerate the diagnostic work: synthesising input, identifying patterns, and surfacing friction points across teams. As the structure started to hold, AI became part of the workflow. Teams used it to generate insights, create materials, and explore ideas faster, within compliance-safe boundaries suited to a regulated environment.
The shift became visible during a three-day working session where teams operated in the new setup on real initiatives. One team created a fully AI-generated presentation video explaining a medical diagnosis in under 30 minutes, something that would typically take days or weeks. The reaction in the room was immediate. Work moved more easily, decisions became clearer, and ownership was easier to see. AI moved into daily use. This is where many initiatives lose momentum, when the system around the work stays the same. Once that system starts to work, both the way of working and the use of AI begin to stick.