GDG on Campus Vienna University of Technology
Modeling chemical reactions is one of the most complex challenges in modern science. For decades, chemists have been lim...
100 RSVP'd
Modeling chemical reactions is one of the most complex challenges in modern science. For decades, chemists have been limited by the slow grind of human trial and error. Today, Deep Learning promises to simulate and predict the behavior of matter at unprecedented speed.
In this talk by Esther Heid, former MIT postdoctoral fellow and leading AI4Chemistry researcher, you'll learn about...
🧪Modeling tasks in chemical reactivity
🦄How to represent chemical reactions for ML-driven property prediction.
✨ State-of-the-art generative AI models (like Flow Matching) to predict transition states.
Whether you are an ML enthusiast looking for high-impact AI applications beyond LLMs, or a natural science student eager to see how Deep Learning is transforming the field, this talk is for you.
📅 Date: March 4th, 2026
🕔 Time: 17:00
📍 Location: GM 3 Vortmann HS
🍕 Perks: Top-tier AI content followed by free pizza and drinks.
Stick around after the talk! Members of the research group will be there to grab a drink and chat about their current work and upcoming projects.
Esther Heid is an Associate Professor at TU Wien bridging AI and chemistry. She received her PhD in Chemistry from the University of Vienna and was awarded the Karl Schlögl-Preis for her work on solvation dynamics. Even though Esther is a chemist by training she cross-specialized into deep learning and became a postdoctoral researcher at MIT. There she co-developed ChemProp, the most widely used Graph Neural Network (GNN) framework for chemistry and invented a new graph representation for chemical reactions. Now, she leads her own interdisciplinary research group at TU Wien working on the next generation of predictive and generative models for chemical reactivity.
Beyond research, Esther is also an active voice in bringing more AI to our university by advocating for modern deep learning courses and leading projects like dataTUdiscovery to connect computer science and data science students with research groups across faculties.
Wednesday, March 4, 2026
4:00 PM – 8:00 PM (UTC)
