At Tune Insight, paths are never linear. Thierry Bossy’s is a great example. Currently a machine learning engineer at Tune Insight, his CV spans research, public health, and field experience far from the digital world. But he has one guiding principle: to make sure that data and technology have a real impact. Meet our data expert who is as driven by understanding complex systems as he is by making them useful, accessible, and practical.
First of all, can you tell us what your role is at Tune Insight today?
I’m a machine learning engineer. In practice, I work on any projects that integrate machine learning and AI, but always with a very strong emphasis on responsibility. My role is to design and develop models, and at the same time to ensure that they can be rolled out and used in a way that is completely compliant with confidentiality and privacy, which are obviously key concerns in healthcare. That also means a lot of discussion with the teams, a lot of testing, and making sure that models work in real life, not just in theory.
If you had to explain your job to someone who doesn’t know anything about machine learning, what would you say?
Did you always know that this is what you wanted to do?
Not really! I grew up in Geneva, and that’s where I went to school. I liked math and physics, I found them easier, so after graduating high school I chose to study engineering at EPFL in Lausanne. What I liked about engineering was that it’s a way to apply math to the real world. That’s what IT was about for me at first: making math useful!
Your path hasn’t been the most direct one…
It hasn’t—thankfully! Between my undergraduate and my postgraduate degrees, I did my national service. At the time, it wasn’t the easiest experience: not much sleep, constant stress, a lot of time outside… But looking back, it was so positive. I was a little older than the others, so not really there to learn discipline. It was more a question of patience. And you really get to know yourself better, and learn to push yourself.
Then there was Ecuador, not a typical destination for an engineer.
Exactly. For part of my national service, I did a non-military placement abroad. I went to Ecuador, near Quito, to work on a recycling project supported by a Swiss association working with an Ecuadorian foundation. The goal was to organize the collection and resale of glass, cardboard, and plastic for recycling. It was very practical. I loved it. The experience gave me an opportunity to take a step back and change how I see things. It didn’t change my career plans, but it definitely enriched my understanding of work and decisions in situations very different to my own.
And then the move to healthcare was a natural one…
Yes, especially given the coronavirus pandemic. For my postgraduate project, I joined a machine learning research laboratory at EPFL, mainly working with Annie Hartley, a specialist in humanitarian medicine. Our project consisted of analyzing public health data from different countries to understand what impacted case numbers and mortality rates: mobility, restrictions, government regulations, weather… Most of all, our work showed how difficult, yet essential, it is to be able to compare data from very different sources. With each country having its own context, and biases like average population age coming into play, you can quickly see how easy it is to jump to the wrong conclusions if data isn’t aligned and used carefully.
And that’s the project that brought you to Tune Insight?
Yes, but indirectly. After that project, it was the research laboratory, in particular Martin Jaggi, who put me in contact with Tune Insight. The idea was to launch a collaborative machine learning project, with a very clear focus on data security and data privacy. It was right in my wheelhouse: working on applied subjects, linked to healthcare, and with major constraints.
What motivates you today at Tune Insight?
Above all, it’s purpose. Health is concrete. Everyone understands what’s at stake. Then, there’s a real challenge surrounding data access. Tune Insight’s tools make it possible to share data in full compliance with ethical principles, legal constraints, and privacy requirements. For me, the company is exactly where it needs to be: at the crossroads between research, healthcare institutions, and data protection.
How would you describe the culture at Tune Insight?
Very collaborative. The team is quite young, and each person has a different, complementary role to play. We’re listened to, and we can suggest ideas. When I arrived, there were seven of us. Now it’s nearly 20. It’s fantastic to see the company keep its original culture as it grows.
And outside of work, you’ve got a not-so-hidden talent…
Yes, I’m a semi-professional salsa dancer. I train all year and compete solo, in a couple, and as part of a group. I learned to dance in Cali, a city in Colombia, where salsa is a way of life. I mainly dance salsa caleña, a very quick and intricate style that’s often danced in competitions.
So what links data, health, and salsa?
I would say… Being creative within constraints, and finding the right degree of freedom to truly innovate. Once you’ve understood the limitations, you know where you can push the boundaries and propose something new—and that’s always a challenge.
