In order to evaluate the climate-related safety of housing locations, this work presents a workable and comprehensive system. The method estimates the impact of three major hazards: drought, extreme heat, and flooding by combining long-term satellite observations with scalable cloud-based processing. We gather and preprocess multisource geospatial data using Google Earth Engine, which are then combined in Google BigQuery to create a single Climate Risk Index. An XGBoost model, which was trained on these features for classifying locations as safe or potentially dangerous, yields a very good performance even on this unbalanced dataset, with an AUC of 0.92. To allow for transparent decision-making, model outputs are explained through SHAP values, pointing out the most important factors behind each prediction. The entire workflow is containerized and deployed on Google Cloud, enabling dependable scaling and fast inference. Risk levels can be explored in interactive maps, including recommendations of neighboring safer areas, through a Next.js web interface. The system also has a conversational assistant that provides concise, intelligible explanations of the results to help users understand the actual importance of risk levels in regard to their area. The platform makes sure users have clear, data-driven suggestions toward climate-ready planning and sustainable city development in line with SDGs 11 and 13.