Abstract: Significant progress has been made in the monitoring and nowcasting technology for severe convective weather, yet the accuracy and lead time still cannot fully meet the needs of disaster mitigation. This article provides an overview of the progress in monitoring and nowcasting techniques for severe convective weather and the development of operational nowcasting systems. It focuses on summarizing the progress of monitoring and nowcasting techniques using deep learning models based on multi-source data, and points out the challenges faced and future outlook. Based on multi-source observations such as those from dual polarization weather radars and geostationary satellites, the monitoring capabilities of severe convective weather types and intensities, convective inception, and identification and tracking of convective storm cells are further improved using structural feature recognition statistics, fuzzy logic, deep learning, and other technologies. The application of deep generative models has significantly improved the accuracy and lead time of severe convective weather nowcasting. The performance of the SWAN (Severe weather Analysis and Nowcasting) 3.0 system of China Meteorological Administration has been significantly improved and it has been widely used in operation in China. In the future, it is necessary to make full use of fine observations and numerical forecasts with the hundred meter resolution, and enhance our understanding of the mechanisms of severe convective weather on meso-γ and microscale. Physics-informed artificial intelligence (AI) models as well as large scale AI models should be developed for nowcasting to continuously improve the monitoring and nowcasting capabilities, eventually realizing the full utilization of forecasters' comprehensive judgment role and enhancing their ability to predict extreme weather.