Impact highlights

From GenAI pilots to scaled platforms

European pharma group · GenAI adoption
70%+ adoption · >3 hours/week saved

Designed and led the rollout of a secure internal GenAI assistant for thousands of employees, with strong governance and education. Heavy users reported 3–4 hours saved per week and the platform seeded multiple vertical AI products (support assistant, knowledge search, patent support).

Industrial manufacturing · Factory efficiency
+4% OEE · >€1.2M/year projected

Co-created a “Factory DNA” framework with operations leaders, using Gen AI on produciton lines data to improve downstream efficiency. Pilots delivered a 4% OEE uplift with a clear playbook to scale across plants.

Grocery retail · Campaign optimization
Hundreds of M€ decisions improved

Introduced a causal-inference approach to marketing and layout initiatives, replacing naive before/after analytics. This changed how a multi-hundred-million-euro promotional budget was planned and evaluated.

Premium manufacturing · Forecasting & maintenance
€4M+ combined effect

Led forecasting and predictive-maintenance initiatives that reduced forecast error and improved inventory and uptime, with direct top- and bottom-line impact.

Stories

How these results were achieved

Three representative stories; more detail and other examples are available on the Stories page.

GenAI, safely, in a conservative group

As a group Chief Data & Analytics leader, I aligned executives, legal and security stakeholders to launch a governed GenAI assistant. We framed it as a capability, not a toy – and used early wins to justify new products.

Read the full story →

Fixing “illusory analytics” in retail

A new store layout was blamed for a revenue drop. Instead of simple before/after views, we applied causal-impact methods and changed how all major initiatives were evaluated – across promotions, pricing and space.

Read the full story →

Factory DNA for pharma plants

Partnered with manufacturing to turn manufacturing steps into a “DNA sequence”, combining Generative AI and data science with practical operator knowledge. The result was a shared language and a scalable blueprint for OEE gains.

Read the full story →