About Me

I am a Machine Learning Researcher specialising in graph generative models, recently completing a PhD in the Learning And Signal Processing (LASP) group at University College London (UCL) advised by Dr. Laura Toni. Prior to joining UCL, I obtained my Bachelor in Electrical and Computer Engineering with a focus on Machine Learning from the American University of Beirut in May 2021.

I work on generative modelling for structured data, with a particular focus on graphs. My work is centered around understanding the trade-offs and limitations of modern generative approaches, including diffusion, flow-based methods, and bridge formulations, and how these interact with underlying data structure. I study how such models capture structural dependencies, and how they can be improved to achieve better controllability, generalisation, scalability, and reliability. I am particularly interested in developing methods that capture both global and local structure, as well as enable controlled transformations between distributions. My work also extends to multimodal generative modelling, where I explore how structured representations can be integrated with other modalities to support more faithful generation and reasoning in complex settings.