The SEWAA project, launched in 2023, uses machine learning, specifically Conditional Generative Adversarial Networks (cGANs), to improve early warning systems for extreme weather in Eastern Africa. By enabling high-resolution, probabilistic rainfall forecasts on personal computers, SEWAA eliminates the need for costly supercomputers, making forecasting accessible and affordable. The project has reduced forecast generation time, lowered costs through cloud computing, and empowered local meteorological agencies with training and ownership of the system. With proven success, SEWAA plans to expand its reach, enhance forecast frequency, and integrate new features, offering a scalable, sustainable model for anticipatory action and disaster risk management globally.