In recent years, deep learning models have rapidly emerged as a standalone alternative to physics-based numerical models for medium-range weather forecasting. Several independent research groups claim to have developed deep-learning weather forecasts which outperform those from state-of-the-art physics-basics models. Thus, operational implementation of data-driven forecasts appears to be drawing near. However, some questions remain on the capabilities of deep learning models to provide robust forecasts of extreme weather. The predictions generated by leading deep learning models appear to be oversmooth, tending to underestimate the magnitude of extreme weather events such as windstorms and cold spells. The aim of this project is to build deep learning models specifically designed for extreme weather situations, bridging the gap in performance between data-driven and physical models. In addition, We want to provide the user with information on variables related to extreme weather prediction, such as maximum wind gust and accumulated precipitation, which are currently overlooked by leading deep learning models. As a first step, we plan to build a regional climate model for the prediction of wintertime extreme weather events in Europe. This model is then to be compared to leading physical and deep learning models in terms of capability of predicting extremes, as well as in terms as of computational requirements and explainability. Given satisfactory performance, the model may in the future be expanded to cover a hemispheric or global scale.