💡 I’m a Postdoctoral Researcher in the Computational Optimization group at the University of Pavia, working with Prof. Stefano Gualandi on MILP algorithms for NVIDIA GPUs and hybrid methods that combine linear programming with deep learning.
🔬 My PhD in Mathematics at Pavia focused on neural operators for PDEs on complex geometries and on neural preconditioners for isogeometric analysis, with research visits at TU Delft and Aarhus University.
🎓 I hold a BSc and MSc in Mathematics from the University of Pavia. My research interests include optimization, scientific computing, scientific machine learning, and the intersection of mathematics and informatics.
🛠️ I like turning research into fast, usable software: HyperNOs, an open-source library for automated, parallel tuning of neural operators, and mateMATTIci Training, a Math Olympiad app used by 1,000+ students and teachers.
🏆 Invited speaker at NeurIPS 2025, 5 journal papers, and 3 Kaggle medals: silver in Google Code Golf (23rd) and NeuroGolf 2026 (40th of 2,963 teams), bronze in Santa 2025.
👨🏫 I teach mathematics and scientific programming at the University of Pavia, including a PhD course on Julia.
💻 In my free time, I enjoy playing volleyball, Valorant, horse riding, and deepening my knowledge of informatics.
✍️ Feel free to reach out for collaborations or discussions on these topics. You can contact me at massimiliano.ghiotto00@gmail.com.
💬 How to reach me:
