Joosep Pata

I am a Senior Research Scientist at the Laboratory of High-Energy and Computational Physics at NICPB, a national lab for fundamental and applied research in Estonia. My work focuses on the intersection of applied machine learning, high-performance computing (HPC), and experimental particle physics.

With over 14 years of experience in scientific computing and deep learning, my research centers on developing scalable machine learning models (such as graph neural networks and transformers) for real-time data reconstruction at the CERN LHC and future colliders. I previously co-led the reconstruction software group for the CERN CMS collaboration and currently direct research initiatives supported by the Estonian Research Council.

Before joining NICPB, I conducted postdoctoral research at Caltech and completed my PhD at ETH Zürich (Thesis). In addition to academic research, I collaborate with industry partners to design and deploy custom machine learning solutions. This includes developing real-time multi-task perception networks for robotics platforms at Taara Robotics, physical AI for 3D tomography at GScan, and sequence modeling at Lingvist. I also advise Mu-Ray Tech.

For details on my background, publications, and talks, please view my resume or get in touch to discuss potential collaborations.