Summary
Valentin Hartmann is an AI research scientist and entrepreneur in Lausanne with a decade of experience at the intersection of machine learning, privacy, and applied mathematics. He is co-founder and CEO of Timely, building language-learning products, while leading research at Xent Labs on novel reinforcement-learning approaches for more generally intelligent LLMs. Valentin completed a PhD at EPFL focused on privacy-preserving ML—publishing and producing practical systems in differential privacy, federated learning and distribution inference—and has a track record of shipping research into production (e.g., a federated SVM deployed with the Cliqz browser). His earlier work includes efficient algorithms for Wasserstein distances that were integrated into the R transport package, reflecting strong numerical and C++ implementation skills alongside formal math training from TUM and Göttingen. He blends rigorous academic methods with startup velocity, often turning theoretical ideas into working infrastructure.
10 years of coding experience
Master's degree Mathematics in Data Science, Master's degree Mathematics in Data Science at Technical University of Munich
Doctor of Philosophy - PhD Computer Science, Doctor of Philosophy - PhD Computer Science at EPFL
Bachelor's degree Mathematics, Bachelor's degree Mathematics at The University of Göttingen
German, English, French