This work addresses assessing air quality and noise in urban environments by integrating predictive models and Internet of Things technologies. For this, a model generated heat maps for PM2.5 and noise levels, incorporating traffic data from open sources for precise contextualization. This approach reveals significant correlations between high pollutant/noise concentrations and their proximity to industrial zones and traffic routes. The predictive models, including convolutional neural networks and decision trees, demonstrated high accuracy in predicting pollution and noise levels, with correlation values such as R2 of 0.93 for PM2.5 and 0.90 for noise. These findings highlight the need to address environmental issues in urban planning comprehensively. Furthermore, the study suggests policies based on the quantitative results, such as implementing low-emission zones and promoting green spaces, to improve urban environmental management. This analysis offers a significant contribution to scientific understanding and practical applicability in the planning and management of urban environments, emphasizing the relevance of an integrated and data-driven approach to inform effective policy decisions in urban environmental management.