About Me
Data science PhD and EPFL graduate with extensive experience developing and deploying data-driven methods for real world applications. Passionate about software engineering and writing.
Timeline
- Building aipr.pub, an AI-assisted scientific peer review platform.
CHUV, UNIL
PhD Student and Postdoc in the Translational Machine
Learning laboratory
October 2021 – February 2026
- Designed and applied computational methods for predicting genetic risk, isolating drug effects, and modeling hospitalization events using large-scale observational real-world data from the UK Biobank
- Data analysis for the HEARTMAGIC study, a collaboration between CHUV and HUG that aims to identify treatment targets for patients with heart failure with preserved ejection fraction
Invision AI
Software Engineer - Intern and Junior
March 2020 – October 2021
- Training and evaluation of deep learning models for cars and pedestrian 3D detection
- Synthetic data generation with Blender Python API
CHUV - Nuclear Medicine Laboratory
Research Assistant - Civil Service
Sept 2017 – Jul 2018, Jul 2019 – Sept 2019, Mar 2021 – Apr
2021
- Helped implementing mathematical models for the assessment of renal function of mice based on PET images
- Assissted in cell culture, radio-chemical binding experiments, article redaction and data analysis
Ecole Polytechnique Federale de Lausanne - EPFL
Master in Computational Science and Engineering
Sept 2019 – Feb 2021
Ecole Polytechnique Federale de Lausanne - EPFL
Bachelor in Physics
Sept 2013 – Jul 2017
Selected Works
GBRS: Gradient Boosted Risk Scoring
Transparent ML library
A gradient boosting approach for generating clinical risk scores. Working
library implemented in C++ with Python and R bindings.
Method and system for extracting clinical markers of cardiovascular health from MRI images
Pending Patent EP20250214061 • Filed 7.11.2025
A novel system for automated extraction of cardiovascular biomarkers from MRI
scans.
Systematic Estimation of Treatment Effect on Hospitalization Risk as a Drug Repurposing Screening Method
World Scientific 2024
A computational framework for identifying drug repurposing candidates from real
world data by estimating their effect on hospitalization risk.