A new beginning

Every blog needs a first post.

Since this is mine, I want to say what brings me here.

Data and AI tools have never been easier to access. A team can turn an idea into a prototype, query its data, or automate a task within hours.

That acceleration creates possibilities. It also shifts the difficulty: what should be built, for whom, and under whose responsibility?

That belief led me to start UpperCase and open this space.

#Value does not disappear, it moves

AI makes it possible to test faster, produce faster, and explore ideas that once required highly specialized teams. But when several solutions become possible, someone still has to decide which one is worth building, in what order, and within which boundaries. Teams also need to know what they would rather not automate.

I encountered this reality well before the recent wave of generative AI. At Galeries Lafayette, connecting more than 20 million customer records on GCP was a technical challenge. Value came from aligning five brands, defining acceptable uses, and turning infrastructure into business decisions.

At Guerlain, a CDP covering more than 18 million contacts across 32 countries required a roadmap, prioritized use cases, data owners, and clear rules for quality and consent.

In both cases, technology expanded what was possible. Value moved toward the ability to choose a direction, understand trade-offs, and build a system that remained useful after the novelty wore off.

#Building with care

A demo is a beginning. A production system has to keep working with real data, real users, and real consequences.

For any data or AI product, practical questions appear quickly. Who is accountable for data quality? What happens when the output is wrong or incomplete? When should a person take over?

These questions shape the architecture, budget, and user experience. They sit at the heart of the Product Owner Data & AI role: connecting business value, technical feasibility, legal constraints, and the team’s actual delivery capacity.

Operational governance prevents a project from moving forward without clear ownership, decision rules, or production criteria.

This is where trade-offs become concrete: a technical choice becomes a product choice, a security rule changes the experience, and an AI promise has to become a set of precise responsibilities.

#Why this blog, why now

My career moved toward this intersection. I built data platforms, then took responsibility for their adoption, roadmaps, and governance. At Guerlain, my role as Product Owner, Data Platform & Privacy brought product, engineering, business, and legal teams together.

Moving from a permanent role to independent work is a way to continue along that path in a more direct setting. I started UpperCase to help organizations lead, structure, or take over data and AI projects that have become too cross-functional for one team to handle alone.

This blog extends that work. I will write about data platforms, AI products, and the governance decisions that determine whether they reach production. Tools will have their place when understanding how they work helps clarify a problem or support a better decision.

I care as much about the foundations as I do about emerging solutions, and especially about what connects them.

I want to write about these subjects in clear language without flattening their complexity. That means keeping the bar high, sharing what works, and looking beyond the noise.

#A clear direction

AI is no longer an isolated topic. It is becoming a force that reshapes the whole system: how organizations design, work, decide, secure, and govern.

Organizations do not need to accumulate tools. They need to select the right uses, set clear boundaries, and bring business, data, IT, and legal teams together to build systems that are coherent, secure, useful, and lasting.

That is where I want to work: where technology meets strategy, innovation meets responsibility, and promises have to become concrete choices.

That is the principle I want to follow here and in my work: understand deeply, explain simply, and build with care.

This first post is not a manifesto. It marks a transition. A way to say where I come from, what I want to keep learning, and where I want to be useful.

What follows will be about AI, of course, but not only. It will mostly be about what has to be built around these technologies for their impact to hold up in the real world.

Welcome.