Summary
Sergey Pozdnyakov is a postdoctoral researcher at EPFL with nine years’ experience at the intersection of geometric deep learning and atomistic modeling. His work focuses on theoretical foundations and practical architectures for 3D point-cloud ML, with notable contributions including PET, ECSE, and rigorous results on the incompleteness of distance-based GNNs and 3-body-based MLIPs. Trained in materials science and data science (PhD at EPFL, MS at Skolkovo, BSc at MIPT), he blends strong physics intuition with advanced machine-learning methods. Based in Lausanne, he leads research in LIAC that bridges provable theory and deployable models for molecular and materials problems. An underappreciated strength is his emphasis on theoretical limits that directly inform more expressive, physically grounded network designs.
9 years of coding experience
2 years of employment as a software developer
Yandex School of Data Analysis
Master's degree, Data Science, Master's degree, Data Science at Skolkovo Institute of Science and Technology
Bachelor's degree, Applied physics and mathematics, Bachelor's degree, Applied physics and mathematics at Moscow Institute of Physics and Technology (State University) (MIPT)
Advanced Educational Scientific Center of Moscow State University (AESC MSU)
Doctor of Philosophy - PhD, Materials Science and Engineering, Doctor of Philosophy - PhD, Materials Science and Engineering at EPFL