Story
Vincennes to Chicago to clinical AI. How I got here.
Vincennes
I grew up in Vincennes, just east of Paris. The kind of place where you walk to school, the boulangerie knows your order, and the Bois de Vincennes is your backyard. I was always drawn to numbers — math competitions, puzzles, the satisfaction of a clean proof.
At fifteen, I left for a year of high school in the United States. That year changed the trajectory. Living in another language, in another system, at an age where everything is forming — it opened a door I didn't know existed.
Chicago
I came to the University of Chicago at eighteen. Six years, three degrees: B.S. and M.S. in Computational and Applied Mathematics, B.A. in Statistics. Dean's List, Susan H. Rudolph Scholarship.
UChicago is relentless in the best way. The culture is intellectual rigor above all else — you learn to think before you learn to build. I studied optimization, signal processing, and probability theory. Research came early: 3D graph signal processing applied to marine biology under Prof. Tingran Gao, then adversarial computer vision and biometric security at the SAND Lab under Prof. Heather Zheng. That work became a CHI 2021 publication.
Chicago is also where I ran my first marathon — October 2019, 4:18:12. No training plan, no idea what I was doing. That became its own story.
Clinical AI
After graduating, I joined Anumana in Boston — a startup spun out of Mayo Clinic, building deep learning algorithms that detect heart disease from standard 12-lead ECGs. The kind of work where the data is messy, the stakes are real, and the validation is as important as the modeling.
At Anumana, I developed an ECG-based algorithm for cardiac amyloidosis detection with Pfizer. That work contributed to an FDA Breakthrough Device Designation. I built NLP pipelines for Mayo Clinic clinical trials, curated multi-site cohorts with Novartis using Spark SQL, and trained CNN and Transformer models on GPU clusters. Published in eClinicalMedicine (The Lancet) in 2023.
In 2024, I moved to Idoven — a cardiac AI company headquartered in Madrid. Led data science on the AstraZeneca partnership: designing clinical data collection frameworks, validating models across European and US hospital sites, presenting at ESC Digital & AI Congress 2025. Published in Heart Rhythm in 2026.
Five years, two companies, four publications, one FDA designation. The thread: using deep learning to find patterns in electrocardiograms that cardiologists can't see with their eyes.
Running
Chicago 2019, 4:18. SF 2023, 3:48. Paris Olympics route 2024, 3:20. NYC 2024 — on a tibial stress fracture, fundraising for prostate cancer — 3:08. Sixty-nine minutes dropped in five years.
I train with RITM in Paris. I've had two stress fractures and kept going. The second one was bilateral — asymptomatic in the right leg, oedeme osseux in the left. Amsterdam 2025 was a DNS. Current focus: Hyrox, triathlon, and eventually Reykjavik.
Film
Three cameras: a Yashica Electro 35 MC, an Olympus Mju-II, and a Rollei Magic II. Mostly film, sometimes digital. I shoot when I travel and when I walk through Paris. The constraint of 36 exposures changes how you see.