The rapid growth of electronic waste (e-waste) poses critical challenges for urban sustainability, requiring advanced predictive tools for effective management. This study develops and validates an artificial neural network (ANN) model to predict e-waste generation in Guayaquil, Ecuador, integrating ambient intelligence (AmI) principles for decision support. Using survey data from 400 residents, we trained an ANN architecture (128–64-1 neurons) that achieved exceptional performance (R2 = 0.912, RMSE = 0.196 tons) in predicting e-waste volumes. Feature importance analysis revealed income level, device acquisition rate, and education as primary drivers of e-waste generation. The model was operationalized through an interactive web dashboard, demonstrating successful implementation of an AmI framework for urban planning. Our approach bridges the gap between predictive analytics and practical decision-making, enabling proactive resource allocation, optimized collection routes, and evidence-based policy development for circular economy implementation in smart cities.