Alessandro Micheli — Imperial College London
I am an AI researcher at Imperial College London working across generative modelling, geometric learning, probabilistic methods, and AI for science.
Alessandro Micheli, Silvia Sapora, Anthea Monod, Samir Bhatt
A framework combining intrinsic entropic optimal transport with amortized neural evaluation on Riemannian manifolds.
Alessandro Micheli, Yueqi Cao, Anthea Monod, Samir Bhatt
A neural optimal transport framework for learning transport maps on Riemannian manifolds.
Alessandro Micheli, ...
A Bayesian deep learning framework for survival analysis with calibrated uncertainty estimates for high-stakes clinical applications.
A seminar on neural optimal transport on manifolds: learning maps between probability distributions while respecting the geometry of curved spaces.
Open on YouTube ↗I'm Alessandro, an AI researcher at Imperial College London. I work on the design of mathematically grounded AI systems, spanning generative models, geometric learning, probabilistic methods, and applications in scientific machine learning.
My route into AI has been through mathematics. I completed a PhD in Mathematics of Random Systems at Imperial College London through the Imperial–Oxford EPSRC Centre for Doctoral Training, and previously studied Part III of the Mathematical Tripos at the University of Cambridge. I also spent time as a quantitative researcher at Virtu Financial, working at the interface of mathematics, statistics, and large-scale data.
My recent work includes papers at NeurIPS 2025 and ICML 2026. Current research directions include neural optimal transport, learning on curved spaces, generative modelling, and probabilistic AI.
