Testimonial - Laura Boutonnet
Data scientistLaura collects, processes and analyses big data to help improve her clients’ performance. A truly pioneering approach!
How did you become a data scientist?
At school, I was good at maths, but I didn’t have a clear career plan. My mum, worried about my future, came across the “BUT – Data Science” programme, and it clicked immediately. It struck the perfect balance between maths and computer science—two core areas in today’s job market. During my studies, I discovered a real passion for data analysis and Python programming. That led me to pursue a Master's degree with the goal of becoming a data scientist.
What skills are essential in your role?
You’ve got to stay on top of developments in data science—it’s a fast-moving field. Being able to explain things clearly to clients is crucial, as is curiosity. We need to fully understand their business processes—whether that’s refrigeration, transport, or agriculture—to offer tailored solutions. And of course, rigour: you need to write reliable code that will be used for years.
What are your other interests?
I’m really into sports—my latest discovery is Hyrox, which helps me unwind after intense workdays. I also share my passion for data through teaching in higher education. Thanks to Equans, I’ve mentored a young person living with a disability, and I occasionally help out with food distribution initiatives.
What a journey since joining Equans as an intern! What drives you today?
I thrive in many ways. On the commercial side, I help clients see the value of their data and how it can boost their performance. I also give a lot of presentations to diverse audiences—engineers, technicians, business managers. I love demystifying AI, coding, and data science. But what I love most is when a client truly understands the connection between their data and how their factory or equipment operates.
Can you give us a few examples?
One of my clients wanted to predict agricultural output—both in terms of quantity and quality. By combining historical data (on yield, quality, past weather conditions) with weather forecasts, we built a model that estimates production for each plot. It was relatively simple but turned out to be extremely useful for the client.
Another project involved developing a programme to predict failures in refrigeration systems. By analysing historical data, I built a model that detects changes in behaviour—flagging the shift from normal operations to early signs of failure. Previously, the client had to react urgently, repair broken equipment and discard spoiled goods. Now, they can anticipate problems and drastically reduce losses.
And now, every time I go shopping, I wonder if their equipment is running on my model—and that makes me proud.
