PRE4IMPACT-AT aims to establish the scientific foundation for impact-based warning (IbW) systems in Austria. The project focuses on the impacts of precipitation-related natural hazards—specifically landslides, flash floods, and hail. Using extensive national datasets and co-design processes with end-users, the project will develop explainable models that integrate meteorological, geomorphological, and socio-economic factors. In addition to short-term warnings, the project investigates long-term trends related to climate change and compound impacts from multiple hazards. By leveraging explainable machine learning (X-ML), PRE4IMPACT-AT contributes to innovative risk reduction, climate adaptation, and the advancement of early warning capabilities in Austria.