Enedis Data Challenge
Time-series imputation for electricity consumption. Implemented Cyclic k-NN and SoftImpute Matrix Completion.
A showcase of engineering solutions ranging from industrial data analytics and software simulation to mechanical design and IoT prototyping.
Enterprise-grade dashboarding solution developed for ArcelorMittal. Replaced legacy Excel workflows with a Python/Streamlit pipeline, enabling real-time monitoring of wire-drawing parameters (tension, temperature, speed).
Time-series imputation for electricity consumption. Implemented Cyclic k-NN and SoftImpute Matrix Completion.
Python-based serious game modeling production chains. Optimized inventory logic and queue management algorithms.
Instrumented frisbee with embedded IMU sensors. Aerodynamic modeling to predict trajectory and optimize stability.
Finite Element Analysis (FEA) of the glenoid component. Analyzed contact mechanics to prevent loosening.
Complete mechanical sizing and material selection (Ashby method) validated by static structural simulations.