Speakers
The workshop features invited keynote and industry spotlight speakers.
Keynote Speakers
Abhimanyu Das
Principal Research Scientist, Google
“Multimodal Time-Series Foundation Models”
Abhimanyu Das is a Principal Research Scientist at Google, where he leads teams on Time Series and Tabular modeling. His current research interests include foundation models for predicting and reasoning over structured data. His team is responsible for the TimesFM and TabFM family of models. He obtained his PhD in Computer Science from the University of Southern California. His research has received multiple paper awards at venues such as ICML, WwW, and WSDM.
Katharina Eggensperger
Associate Professor, TU Dortmund University
“Scaling and Understanding Models for (Scientific) Tabular Data”
Katharina Eggensperger is an associate professor for ML and AI at the Lamarr Institute and TU Dortmund University. Before that, she led an early-career research group in the Cluster of Excellence Machine Learning for Science at the University of Tübingen. She received her PhD from the University of Freiburg, under the supervision of Frank Hutter and Marius Lindauer. Her research focuses on automated machine learning for tabular data, with the goal of advancing both AutoML methods and foundation models to better support applications in science. She has co-developed widely used open-source tools for AutoML and hyperparameter optimization. Her work emphasizes rigorous empirical evaluation and benchmarking to advance robust, reproducible machine learning research.
David Holzmüller
Research Scientist, INRIA
“TabICLv2: Advancing Open Tabular Foundation Models”
David Holzmüller is a Paris-based researcher at INRIA, working on machine learning for tabular data and uncertainty quantification in collaboration with the groups of Gaël Varoquaux and Francis Bach. His research spans tabular foundation models (TabICL), deep learning models (RealMLP), benchmarks (TabArena), and more. Previously, he obtained his PhD from the University of Stuttgart under the supervision of Ingo Steinwart, exploring topics such as active learning, neural network theory, neural networks for atomistic simulations, and sampling.
Industry Spotlights
Amazon AWS — Oleksandr Shchur
Sr. Applied Scientist
“Chronos-2: From Univariate to Universal Forecasting”
Oleksandr Shchur is a Senior Applied Scientist at AWS, where he develops open source forecasting tools such as AutoGluon-TimeSeries and Chronos.
SAP — Sam Thelin
Principal Scientist
“The Idiosyncrasies of Enterprise Data – the SAP Experience”
Sam Thelin is a principal data scientist at SAP. Before becoming AI pilled, he was a research mathematician, receiving his doctorate from Oxford University.
Nixtla — Max Mergenthaler and Cristian Challu
Co-Founders
“What Is a Good Forecast? Reflections on Accuracy, Architectures, and Incentives”
TimeCopilot — Azul Garza and Renée Rosillo
Co-Founders
“Agentic Forecasting: Scaling Time Series Foundation Models in Practice”
Azul Garza 🇲🇽🏳️⚧️ is an AI and Machine Learning leader with deep experience in econometrics and mathematical modeling. Co-Founder of TimeCopilot and co-creator of the Nixtlaverse and TimeGPT. Former CTO and Co-Founder of Nixtla. A frequent live-coding speaker at ISF, Microsoft Build, AWS re:Invent, and PyCons. Published researcher and recipient of multiple forecasting and AI awards.
Renée Rosillo is a technologist and entrepreneur working at the intersection of AI and inclusive innovation. Co-Founder of TimeCopilot. Renée also co-founded AO Labs, working on continuous learning weightless neural networks. Participant of Creative Destruction Lab Toronto AI at University of Toronto. She previously co-founded Prism, a global network for queer founders, and co-founded Minerva Robotics and FounderFamilia.
TimeCopilot is the universal time series forecasting framework.
Fundamental — Kevin Scaman
ML Researcher
“What LLMs Learn (and Don’t) from Tables”
Kevin Scaman is part of the research team at Fundamental, an AI company pioneering the future of enterprise decision-making. Their most powerful Large Tabular Model, NEXUS, is purpose-built for the structured data that drives enterprise decisions. Before joining Fundamental, Kevin was a research scientist at Inria Paris and a part-time Associate Professor at École Polytechnique. His research in machine learning spans from theoretical advances in key aspects of deep learning, including robustness, decentralized learning, and training via non-convex optimization, to practical implementations of graph neural networks extending their expressive power and robustness.
StableAI — Xingxuan Zhang
CTO, LimiX Project Lead
“Unleashing Structured-Data Modeling Capability for Generalist Intelligence”
