SHORT BIO
Hi, I’m Nadia — a first-year Ph.D. student in Computer Science at the Process & Data Intelligence Research Unit of Fondazione Bruno Kessler (FBK), and the Free University of Bozen-Bolzano (UniBZ), supervised by Massimiliano Ronzani (FBK) and Chiara Ghidini (UniBZ).
My research lies where causality and representation learning intersect, viewed as tools for building trustworthy AI systems. I’m intrigued by out-of-distribution scenarios, where things get messy, assumptions fall apart, and models reveal their true personality when faced with the real world.
I earned both my Bachelor’s degree in Information and Business Engineering and my Master’s degree in Artificial Intelligence Systems from the University of Trento. Along the way, I took part in the Industrial AI Challenge with BLM Adige, where my team won the competition and translated our work into publications on demand forecasting for supply chain management, in collaboration with Marco Formentini. I later joined the U2IS Lab at ENSTA Paris as a research intern, working under the supervision of Gianni Franchi on representation diversity for out-of-distribution generalization in time series data. There, a spark for latent representations took shape and grew into the foundation of my Master’s thesis, co-supervised by Stefano Teso.
Now, I’m still snooping around latent spaces, trustworthiness, and causality — sketching the sweet spot where theory meets reality, where models learn to stay meaningfully grounded.
INTERESTS
- Representation Learning
- Causal Inference
- Out-Of-Distribution Scenarios
- Trustworthy AI
- Recommender Systems