
Jan. 2026 -- May 2026
Evident Scientific
AI R&D Software Engineer Intern
At Evident Scientific in Québec, I worked on an end-to-end module for evaluating AI models dedicated to segmentation and anomaly detection, integrated into a large Python desktop application.
- Designed a hybrid architecture combining PySide6, Tabulator, and JavaScript to interactively visualize thousands of inference results.
- Optimized RAM and VRAM usage during batch inference.
- Reduced storage footprint by 90% through sparse matrix compression.
- Implemented advanced metrics for defect detection evaluation on wind turbine blades.
This internship strengthened my experience in applied AI tooling, desktop software architecture, and performance-conscious data visualization.


