Founding Research Engineer at Noumenal Labs
physical AI · probabilistic modelling · physics
I'm a physicist and machine learning engineer. ML has been the common thread throughout my career. I've applied it to many-body physics, quantum computing, protein design, thermodynamic computing, and scientific discovery with agentic AI.
At Noumenal Labs, I work on physical AI: building systems that bring together probabilistic modelling, physics, and AI to create robots that understand and adapt to the real world.
I started as an electronics technician in Switzerland. I did a four-year apprenticeship starting at 16, then worked in the field while studying physics on the side. That hands-on hardware background still shapes how I think about computation.
I studied physics at the University of Basel (MSc, 2014–2016), then moved to Barcelona for my PhD at ICFO (2017–2021) as a Marie Curie fellow. My thesis was on using machine learning to study many-body quantum systems, training neural networks to discover phase transitions and identify new order parameters in disordered systems.
From 2021 to 2022 I did a joint postdoc with IBM Research and EPFL's Computational Quantum Science Lab, moving deeper into quantum machine learning: building quantum-enhanced ML models and using classical ML to optimize quantum circuits. This included collaborations with Xanadu and Google Quantum AI, and winning the Qiskit Hackathon Switzerland for integrating Qiskit into PyTorch's autograd pipeline.
I then spent 2022 at Menten AI, applying quantum computing and quantum-inspired algorithms to protein design, an exciting intersection of computational biology and quantum optimization.
In late 2022 I joined Extropic as a Staff Scientist, architecting the first core thermodynamic inference systems: physical hardware that performs probabilistic inference through natural dynamics. I built platforms fusing energy-based models with thermodynamic chips.
I then spent about a year at Axiomatic as AI Lead Scientist, building LLM-based agentic AI systems applied to physics research in photonics, accelerating the design and fabrication of photonic devices.
Now at Noumenal Labs, I'm back where I'm most at home: the intersection of probabilistic modelling and AI, building thermodynamic brains for robots with probabilistic ML.
Full list on Google Scholar.
linkedin · github · x/twitter · google scholar
Barcelona, Spain