We look forward to welcoming Lucie Flek, Wolf Vollprecht, and Thomas Kluyver as keynote speakers at NOBUGS2026.
Prof. Lucie Flek
Lucie Flek is a professor at the University of Bonn, where she leads the Data Science and Language Technologies group. Her research centers on machine learning for natural language processing, focusing on AI safety, robustness, and applications ranging from large language models and dialogue systems to clinical text analysis, mental health, misinformation, and social media. With experience spanning both academia and industry, she has led language understanding projects at Amazon Alexa and helped launch Google Shopping Search in Europe. She earned her PhD at TU Darmstadt on tackling meaning ambiguity in deep learning and has held research roles at the University of Pennsylvania and University College London. Earlier in her career, she even contributed to particle physics research at CERN, studying axion searches.
Thomas Kluyver
Thomas started his career as a plant biologist, but during his PhD he began to contribute to open source software. Then a chance came up to do a postdoc working on the IPython project, and what was soon to become the Jupyter Notebook. Working on programming full time, he has created and contributed to a wide range of projects, from scientific libraries to packaging tools, command line utilities and GUIs. Now employed at European XFEL, he is, among other things, one of the main maintainers of h5py, a widely used Python library for working with HDF5 files.
Wolf Vollprecht
Wolf is a core contributor to the conda/conda-forge ecosystem. He studied robotics at ETH Zurich and started his career in open source scientific computing at QuantStack in Paris, where he worked on the xtensor numerical library, the mamba package manager, and improvements to Project Jupyter. He also created RoboStack, bringing the ROS robotics framework to conda-forge. He founded prefix.dev in Berlin, the company behind the Pixi package manager and rattler-build, with the mission of making software packaging reproducible, secure, and cross-platform. Pixi is used across scientific disciplines from bioinformatics to high energy physics.
