Flooding is an increasingly destructive hazard globally, and timely and accurate prediction is essential for issuing early warnings, guiding infrastructure planning, and supporting emergency response—ultimately saving lives and reducing economic losses. During I-GUIDE 2025 Summer School, we explored the development of an advanced flood prediction framework using machine learning—particularly multimodal deep learning—to fuse multiple publicly available geospatial datasets. We will describe the training models that integrate remote sensing, environmental, and ground-based data, and how we assessed how data modality contributes to overall model performance to better understand inter-modal synergies.