Experience
Senior Data Scientist
Grenoble (FR)Machine learning physical dynamics (Scientific Machine Learning) on High-Performance Computers (HPCs). Atos Big Data & Security - Ai4Sim team
-
Research in AI for Physics
Research on machine learning fluid dynamics with Hamiltonian Neural Networks (HNNs), Physics-Informed Neural Networks (PINNs), Auto-encoders. Applications to weather forecasting and chaotic dynamical systems.
-
R&D collaboration with the Airbus Flight Physics department
Surrogating complex numerical simulations of fluids around airfoils for the aeronautic industry. Development and tuning of convolutional deep learning models on HPC platforms. Coupling between CFD solvers and machine learning graphs.
-
Member of the Atos Expert Community
Computer Science Engineer
Grenoble (FR)-
Fullstack developer
Develop and maintain a full and high-disponibility data-driven service. Embedded SDK, frontend dashboards, servlets, backend computing cluster.
-
Data Scientist & Engineer
Digital signal processing and machine learning stack to provide context-aware, high-level semantic informations. Decision making algorithms in the challenging environment of embedded systems. Robust data pipelines and algorithms with a variable number of sensors, data sources.
Engineering school internships
Walldorf (DE)-
Master thesis student
Mar 2016 - Sep 2016Learning to rank strategies of investment. Training algorithms with financial data to optimize investment strategies made of bonds & equities with respect to different risk and performance measures. SAP HANA Vora team [PDF]
-
Second year internship
Jun 2015 - Sep 2015Ranking aggregation for portfolio optimization. Use of different risk-driven performance measures, eg. Sharpe ratio, Value at Risk, Maximum drawdown... Strong emphasis on stochastic optimization heuristics like the Cross-Entropy Monte Carlo method for large-scale and complex problems. SAP HANA Vora team
Research Assistant
Grenoble (FR)Graphical Bayesian models for high-dimension regression using Gaussian mixtures. Applied to the retrieval of physical properties from hyperspectral images. Modelling spatial dependencies with Markov Random Fields. INRIA MISTIS team
Software Developer
Grenoble (FR)Simulating the behaviour of a gas circuit breaker (protecting electrical power stations) for a major industrial group. Development of a pre-dimensioning tool (low-level optimized C-language). Numerical methods for solving mechanical systems, fluid dynamics and thermodynamics equations.
Education
Research-oriented Master
Grenoble (FR)Applied statistics, Machine Learning and Data Science, Convex optimization, Stochastic calculus for finance. Grenoble INP, IM²AG - MSIAM program (Master of Science in Industrial and Applied Mathematics )
Engineering school
Grenoble (FR)Two years with a strong emphasis on software development (C/C++, Java, functional languages, MapReduce / Big Data etc.) and applied mathematics / statistics and their usual tools (Python, Matlab, R). Advanced algorithmics (complexity theory, AI projects). Mathematical Modeling, Image & Simulation
Preparatory classes
Toulouse (FR)Two-year intensive program preparing for the national competitive exam for engineering schools (Grandes Écoles). Major in mathematics, minor in physics and engineering. MPSI/MP
Personal interests
I practice mountaineering and martial arts. I have a specific interest for history, cultures and languages. I speak fluent French and English. I also have a special interest for linguistics, especially the Germanic and Romance language families.