AFAM:
Hello, Louis, and thank you for agreeing to answer my questions. You graduated from both Arts et Métiers and UC Berkeley (Electrical Engineering and Computer Science). Since 2019 you have been working in the healthcare industry as Machine Learning Engineer: first at VideaHealth, an AI platform for dentists and today at Butterfly Network, an ultrasound company. Please tell us a few words about your mission with these companies and whether working in the healthcare industry was a coincidence or your choice.
Louis:
Hello! Thank you for having me. Working in the healthcare industry was definitely a deliberate choice for me. I've always been interested in the intersection of technology and healthcare, and I wanted to work at the crossroad of AI research and AI products.
During my time at VideaHealth, my mission was to develop and deploy machine learning algorithms to automate the analysis of dental X-rays and assist dentists in their diagnosis. The goal was to improve the accuracy and efficiency of dental diagnoses, ultimately leading to better patient care. The main project I worked on was developing an algorithm to detect cavities on X-rays and get that cleared by the FDA. Being able to work on solving such a problem once and then being able to deploy it to thousands of dentists to help the diagnosis of millions of patients is truly what is amazing about working in ML.
Butterfly Network has developed a handheld ultrasound device called the Butterfly iQ. My mission here is to develop algorithms to run on top of the imaging capabilities of the Butterfly iQ. For instance, displaying in real-time the quality of the ultrasound image or labeling anatomical features.
AFAM:
How do you think working in the healthcare industry is different from working in other companies?
Louis:
The main difference is that the algorithms I develop can bring a lot of benefits to patients as long as they are carefully built. That is why the FDA makes sure all of those algorithms are fair. This has direct implications on the code I write since there are important requirements in place that are meant to safeguard against mistakes.
Another key aspect is the cross-functionality of the work. As a Machine Learning Engineer, I work with people in the Software, Product, Clinical, and Regulatory teams. As such, being able to understand and work with them is key. In this regard, I believe the French Grandes Ecoles system and Arts & Métiers definitely helped in building a broad set of skills.
AFAM:
Boston in the US is considered to be one of the centers of biomedical engineering, research, and innovation. Do you take advantage of being in Boston to advance your career in the biomedical sector?
Louis:
Absolutely, a great example of that would be the MIT Healthcare hackathon that is happening every year and that I got the chance to win last year. Boston also has so many hospitals that it makes it very simple to be in touch with the latest research.
AFAM:
Any biomedical startups/companies in the US or in France you are following or/and have been impressed by recently? Why?
Louis:
There are quite a few! On top of my mind, I would mention Digital Diagnostics, a company that works in the ophthalmology space, which is one of the rare companies that got an FDA clearance for an AI algorithm that can be used completely autonomously, which allows it to diagnose patients in areas where ophthalmologists are not available.
There is also Medivis, a company working on imaging for surgeries, and we should see significant breakthroughs from them soon!
In France, Cardiologs is doing an amazing job in the ECG space and while being a French company already has an FDA clearance!
AFAM: in our Shasta program, we often help students interested in the healthcare industry and also those interested in Machine Learning. What would you advise these students who are willing to find an internship in the US?
Louis:
In terms of key skills to master in order to get hired in such positions I would mention coding and data skills. It’s necessary to be great at ML algorithms but without solid skills in Python and leveraging the pandas library it is very difficult to work efficiently.
AFAM:
Any future plans concerning your professional career you’d like to share with us?
Louis:
The coming years will definitely be quite challenging based on projects I cannot disclose yet but I am hoping I’ll deliver amazing news in a year!
AFAM:
Thank you for your time, Louis, and have a wonderful day!
Other interviews in our "Medi, Vidi, Vici" interview series:
Interview with Alexandre Becache (Bo 219)
BLOG / 
CALENDAR / 
DONATIONS / 
MENTORS /