Lukas Fischer

In a world of ever-growing AI models, a small but mighty time series foundation model from Austria is challenging the rules. TiRex, an xLSTM-based model developed by NXAI, has shown competitive zero-shot forecasting results alongside models from Amazon, Google, Salesforce, and Alibaba—despite using far fewer parameters and less computational power. In this exclusive TEDAI workshop, Lukas Fischer, Head of Applied Research at NXAI, takes you on a fast-paced journey into the heart of time series innovation. We’ll crack open the architecture of xLSTM, unpack why TiRex works so well in zero-shot settings, and then jump straight into live forecasting. Bring your own CSVs—yes, your real-world data—and put TiRex to the test on your laptop. Expect a hands-on session where predictions take center stage.Whether you’re a data scientist, ML engineer, or just a curious tinkerer wondering if foundation models have a future beyond language, this workshop is your invitation to forecast the future—literally. Laptops required. Some familiarity with data formats (e.g., CSV) is helpful, but no deep learning experience necessary.