Why a career in tech?
I had a certain facility with mathematics and so naturally enough, I engaged in a scientific career. This allowed me to develop a natural curiosity for technological advances. Tech is an exciting field that offers research and development opportunities in many different application areas. I am particularly interested in its ability to interconnect different fields and offer innovative solutions to current and future challenges.
Your professional experience?
After obtaining my "baccalaureate" in Brest, I wanted to continue my studies in mathematics, but I did not want to pursue "classes préparatoires" because of the emphasis on physics (I was not interested in physics at the time and I thought that "classes préparatoires" was just math/physics) and the competition between students. As an alternative, I opted for a Bachelor's degree in Applied Mathematics for Social Sciences in Brest. This experience was enriching because I could explore the beauty of mathematics while studying diverse fields like economics. Subsequently, I pursued a Master's degree in Applied Statistics for Business in Rennes, where I developed a passion for statistics. After completing my Master's degree, I was unsure of what career path to follow. However, I secured a job as a research engineer in the statistics laboratory of an agronomy school, enabling me to pursue a PhD thesis simultaneously.
My experience of the statistical research world was a revelation, and I was recognized for my thesis as the recipient of the prize for the best thesis in applied statistics from the Société Française de Statistique. I also received a Marie-Curie scholarship that enabled me to spend 18 months at the Department of Statistics at Stanford University. These experiences opened up several opportunities for me, including a professorship at Ecole Polytechnique and a visiting researcher position at Google Brain. I am currently a researcher at Inria, where I founded the Inria-Inserm Premedical team. Our research focus is on precision medicine through data integration and causal learning.
Your first experience with technology?
My doctoral work, where I had to organize wine tasting experiments and develop methods to analyze the resulting data - Yes, we can do many different things with tech!
What do you do today, and why?
As the head of a team of researchers in statistics, machine learning, and clinicians, I am involved in the development of digital health methods for personalized medicine. Specifically, I am interested in the issue of "generalization", which involves predicting the effect of a treatment on a new population based on the results of a clinical trial conducted either in the past or on a different population. The societal implications of these techniques are significant, as they would enable faster introduction of treatments into the market and address the limitations of clinical trials, such as their low patient representation and high cost. There are also economic implications, as the price of drugs can depend, among other things, on their effectiveness.
Beyond their application, the development of such techniques comes with methodological challenges in causal inference, such as managing changes in data distributions and handling missing data when multiple sources of information are available. My passion for this field stems from the opportunity to find solutions to these challenges by combining advances in statistics, AI, and medicine to improve public health.
Your strengths in this role?
I have a certain interest in projects with social and interdisciplinary vocations. Moreover, I have been exchanging with clinicians on research projects for over 6 years, so I am more efficient at quickly understanding them, understanding the data, and associated issues.
Finally, although I remain inherently passionate about research developments, I also enjoy managing a team, transferring my skills, and supervising students.
Past challenges, failures and disappointments?
My approach to research is at the intersection of theory and application, and this position can sometimes be difficult. My goal is to develop statistical and machine learning methods that meet concrete needs in many application domains while being firmly grounded in solid theoretical foundations. This is essential to understand their regime of predilection and to know how to use them as safely as possible. To achieve this, I often collaborate with teams that have complementary skills to cover all needs.
I was pained by results and comments that I considered unfair in the context of project calls. For example, I submitted an application for a chair that I was particularly proud of and on which I had invested a lot. The rejection comment stated that my health applications were only pretexts and that they had purely illustrative value. This was difficult to accept given that I work closely with physicians from the beginning of projects and that I am committed to implementing the methods developed until the patients' beds.
Best moments, successes you’re proud of?
Fortunately, there are many good moments.
Since 2016, I have been working with the Traumabase group to improve the management of polytraumatized patients (victims of road accidents, falls, stab wounds), who represent the second leading cause of mortality and disability in young adults. We are developing decision support models to better anticipate the necessary resources (blood products, control of bleeding: surgery, interventional radiology, etc.) from the initial care by the SAMU (French emergency medical service) to the hospital and to better transfer patients to specialized centers. The goal is to speed up therapeutic management, which is crucial to reduce mortality and improve the functional outcome of patients. It should be noted that the database, which is key to developing these models, was created as a result of initiatives by clinicians aware of the difficulty of decision-making in an uncertain, stressful, dynamic environment with many stakeholders and information fragmentation. I am delighted to have been able to help find funding and improve data collection and analysis. I am particularly enthusiastic because we are currently testing the Shockmatrix mobile application in real-time in ambulances, which aims to predict the occurrence of hemorrhagic shock, and we plan to conduct a randomized trial to evaluate the benefits of such an application.
I also have very fond memories of my research stays at Stanford, which enriched me both personally and scientifically. More recently, I stayed at the Simons Institute in Berkeley, which is conducive to scientific breakthroughs by bringing together many researchers in ideal conditions and bringing different communities closer together.
Furthermore, I also have very good moments during stimulating scientific events such as the UseR! conference for R statistical software users, which brings together researchers from different disciplines around the R statistical software. Beyond the scientific presentations, the social events are really conducive to beautiful encounters and exchanges.
Finally, a comment from a student or a person at the end of a course or presentation satisfies me fully.
People who helped, influenced -or made your life difficult?
First of all, my teacher in the linear regression course in Master's degree, Eric-Matzner Lober, sincerely gave me the passion for the field and directed me to my first job at the Agronomy school. Then François Husson, my thesis supervisor and permanent collaborator, always supported me throughout my career. Finally, at Stanford, Trevor Hastie, Susan Holmes, Persi Diaconis, John Chambers, and Naras Balasubramaniam warmly welcomed, supported, and greatly inspired me. And of course, my family.
Your hopes and future challenges?
Currently, I am interested in even more ambitious projects, with a short-term impact on society. I am particularly excited about a recent collaboration with the Idesp (Inserm-University of Montpellier Research Unit) on the exposome, particularly regarding respiratory diseases such as asthma. The available data is exceptional, with epigenetic, microbiome, environmental, dietary, and socioeconomic information, among others. The statistical and causal challenges for better understanding the effects are considerable, and the potential impacts for improving prevention are significant.
I am also eager to encourage more interactions between the academic world, businesses, and the government. In the context of my work, I have developed causal inference methods that could prove very useful in evaluating treatment effects. For this, it is essential to work closely with health authorities. In addition, I have always wanted to put my skills at the service of humanitarian organizations.
One of the most enjoyable aspects of being a researcher is that you are constantly learning, discovering, and imagining new things, as well as tackling stimulating and interesting projects.
What do you do when you don’t work?
Outside of work, I devote most of my time to my 2-year-old daughter and my family by going on small hikes, reading stories, and so on. I also have a strong interest in traveling because I grew up in several countries and part of my family is abroad, but I now try to travel more responsibly. To fulfill my passion for travel, I read books by travel journalists and writers such as Kapuściński and Tesson. I also enjoy listening to the kora, an African musical instrument.
I am passionate about science, particularly natural sciences. Therefore, I read scientific books and articles from journals such as Science or Nature, and I listen to scientific programs on topics such as the wood wide web, cephalopods, spiders, etc. I would also like to observe more birds and read more astrophysics.
Finally, I enjoy attending concerts and going to the cinema, even though I no longer have as much time for it.
Your heroes -from History or fiction?
I don't have particular heroes, but I am inspired by many figures and collectives, even though there are certain parts of their history that I do not necessarily adhere to. The first names that come to mind are those of Charles Darwin, Hedy Lamarr, Alan Turing, Srinivasa Ramanujan, Marie Curie, Rosalind Franklin, and the entire team that worked with her, Gregor Mendel, and Jean-François Champollion, but there are so many others.
In general, I am fascinated by the multidisciplinary aspect of some scientists who were both mathematicians, philosophers, astronomers, and more. Finally, I think it's fantastic that Thomas Pesquet inspires so many people.
A saying or proverb you like in particular?
I am a positive person, so I often use the expression "it can't get any better" to express my satisfaction at work as well as in my personal life. Of course, as a researcher, there are always new challenges and tasks to accomplish, but I find it important to celebrate each milestone and work with pleasure. For me, fun and passion are key elements to maintain my long-term motivation and commitment.
A book to take with you on a desert island?
Something big ...as a universal encyclopedia.
A message to young female professionals?
I encourage you to consider a scientific career! Mathematics is at the heart of many fields of application and is necessary throughout society. The profession of researcher is exciting and offers the opportunity to discover new horizons. However, it is important to carefully choose your supervisors at the beginning of your career to feel encouraged and avoid disappointment. Once established, this profession offers the freedom to choose your collaborators, your projects and much more. As my colleague Claire Boyer says, knowledge sets you free.
Please feel free to contact me or any of the PhD students to learn more about this profession.
