I was born in Paris and grew up on the Caribbean island of Guadeloupe and initially came to the US on a college basketball scholarship. My passion for numbers and building models led me to a master’s programs in data science. After working in New York and North Carolina for a few years, I decided to take my life and career back to Europe. McKinsey then offered me a role in Amsterdam in 2019, where I am now a principal data science lead at QuantumBlack, AI by McKinsey.
My role is all about helping clients design, deploy, and scale AI solutions—whether that means optimizing production at a mine, improving efficiency at a factory, or helping companies navigate the next wave of generative AI. Throughout the years I have witnessed how AI can impact governments and industrial giants alike, less mature companies as well as the already technologically advanced. Every project is different, but the goal is always the same: to create impact that last while bringing people along in the process.
Mining for insights in unexpected places
Some of my favorite work has been in the mining industry. Imagine a site with hundreds of sensors tracking data points like water flows, densities, temperature, and metallurgical composition.
Traditionally, operators often relied on fixed setpoints and rules of thumb developed over years of experience—running the plant “the way it has always worked.” My team and I used data to bring science to that intuition. We built predictive models that showed how changing one variable, like water flow or pressure, would affect yield and quality. Those models helped operators make real-time, data-backed decisions that improved both efficiency and output. More than the numbers, it also impacts the way people think and approach their day-to-day operations.
It’s incredibly satisfying work. You can literally see the results coming down the other end of the line. There’s no hiding behind theoretical numbers; the impact is very tangible
Boots on the ground
I spend time on-site with clients, sometimes for weeks or months at a time. Whether at a mine, chemical plant, or mayonnaise factory, I’ve enjoyed being out there in boots, covered in dust, standing next to the machinery my models are helping to optimize.
When I see a parameter like “water flow” on my screen, I can walk outside and see the actual pipe it refers to. That makes the data come alive. It also makes the results very real—you know right away if what you’ve built works or not. Either you’re producing more iron or coal, or you aren’t.
I love the people in those industries too. They’re pragmatic, direct, and focused on results. It’s not about flashy technology; it’s about whether it works. Staying in a mining camp in a remote part of Australia, eating at their cafeteria, taking the bus at 5:00 a.m. to the mine, I get to experience their rhythm and day-to-day schedule.
Collaboration at scale
What’s unique about McKinsey—and QuantumBlack—is how multidisciplinary the work is. On one project at a chemical plant in France, we were trying to understand why two identical reactors were performing differently.
That kind of collaboration—bringing together strategy consultants, engineers, chemists, and data scientists—is what makes McKinsey special.
We printed out every part of the process and covered the walls with diagrams. In one room, mechanical engineers walked through every pipe and valve. In another, chemical experts debated reactions. The data science team was analyzing hypotheses and patterns from the models.
By combining all that expertise, we found subtle process differences that explained the gap. That kind of collaboration—bringing together strategy consultants, engineers, chemists, and data scientists—is what makes McKinsey special. It’s a team sport.
Building the future of AI
These days, I’m working on some of the newest frontiers in AI, like agentic systems that combine generative models with real-time data and enterprise integration. I’m currently part of a team leading a multi-month project to design and build an agentic platform that helps our client develop and scale Voice AI use cases.
Technology is moving incredibly fast—new models and frameworks come out every week. Part of my role is to help clients cut through the noise and figure out what really matters for them. Not every new tool is right for every situation, and sometimes the smartest move is to wait.
Still, it’s an exciting time to be in this space. Every day we’re learning something new, and it feels like we’re helping to shape the next generation of how organizations use AI.
Whether I’m building AI platforms or watching trucks deliver beets to the factory, my motivation is the same: to make data meaningful.
At its best, data science is about people—helping them make better decisions, do their jobs more efficiently, and achieve things they couldn’t before. When I see that happening, when I see impact in action, that’s when I know we’re doing something truly powerful.
Related
