The flood susceptibility map serves as a crucial resource for policymakers and authorities, providing a spatial understanding of vulnerability. The research aims to create a flood susceptibility map for the lower Vellar basin, utilizing seven key factors: Curvature, Dem, Drainage Density, Slope, SPI, TWI, and LULC. In this study, Geographic Information System (GIS) and Artificial Neural Networks (ANNs) were integrated to map the flood susceptibility. A total of 150 samples, representing flood and non-flood conditions, were collected from various locations. The dataset underwent a division into 75% for training purposes and 25% for testing. Subsequently, an artificial neural network was employed to construct a flood susceptibility model. The findings reveal that 25.73% of the study area is categorized as being in the very high susceptibility zone, with an additional 24.04% classified as highly susceptible. Moreover, 9.18%, 17.98%, and 23.07% of the area demonstrate very low, low, and moderate flood susceptibility, respectively. The validation process highlights an impressive overall prediction accuracy of around 86.3%. By supporting resilient infrastructure (SDG 9), sustainable cities (SDG 11), and climate action (SDG 13), flood susceptibility mapping contributes to achieving the Sustainable Development Goals (SDGs). These findings provide a valuable spatial framework for sustainable land use planning, disaster preparedness, and resilience-building in flood-prone landscapes.