About me
I am an AI Engineer in the Privacy-Preserving Machine Learning and Vision Foundation Model team at Sony AI Zurich, where I design, develop, and scale cutting-edge vision and multimodal foundation models that power real-world applications across Sony’s ecosystem. My work spans the entire lifecycle of model development—from data curation, model architecture design, and efficient training, to evaluation, performance tuning, and deployment—with a strong emphasis on scalability, efficiency, and safety.
Before joining Sony, I was a research assistant in the Social Computing group at Idiap Research Institute in Martigny, Switzerland, where I worked on machine learning for social systems and privacy-aware applications.
I obtained my PhD from EPFL, working on privacy-preserving machine learning on graphs. During my PhD, I did an internship at Brave and visited the Safe and Ethical AI programme at The Alan Turing Institute.
My research interests lie at the intersection of trustworthy machine learning, computer vision, and multimodal foundation models. I am particularly interested in building robust, efficient, and privacy-aware AI systems that can be reliably deployed in real-world applications.
You can find more about me in my CV.
News
Our paper StelLA: Subspace Learning in Low-rank Adaptation using Stiefel Manifold has been accepted to NeurIPS 2025 as a spotlight!
Our paper “Argus: A Compact and Versatile Foundation Model for Vision” has been accepted to CVPR 2025.
Our paper “ProGAP: Progressive Graph Neural Networks with Differential Privacy Guarantees” has been accepted to WSDM 2024. The code is available on GitHub.