It is beyond contestation that a sound data ecosystem can help drive strategic decisions towards building disaster-resilient infrastructure, cities, and societies. However, data that could enable such strategic decisions is scattered across different agencies, scales and formats, limiting their utility for informed decision- making. This results in inefficient actions or ad-hoc responses that fail to cater to urgent needs during emergencies or provide sustainable long-term solutions. There is an urgent need to commit towards a comprehensive G20 Climate Data Ecosystem that allows modelling intelligent solutions using advanced artificial intelligence and machine learning techniques to inform decision-making. Such an ecosystem needs to look at climate data holistically and make interoperable, AI-ready, standardised datasets available across current data silos.