Pneumonia is an acute respiratory infection leading to morbidity and mortality globally. Timely and correct diagnosis would reduce its impact, especially in seasonal outbreaks and pandemics. Conventional methods of diagnosis rely on the manual interpretation of chest X-ray images; they are time-consuming and have variability, which can hinder rapid public health responses. This paper proposes a CNN-based framework for automatically detecting pneumonia from chest X-ray images, targeting accuracy in diagnosis and scalability toward large-scale epidemiological surveillance. The CNN model achieved an excellent accuracy of 92.5% when using robust data in both the training and validation stages to classify normal images versus pneumonia-affected images. This accuracy level indicates the model’s potential as a diagnostic aid, offering near-instantaneous results that can be integrated into public health frameworks for real-time disease surveillance. Outputs of the model will have much broader epidemiological applications because they can support public health efforts aimed at tracking infection patterns, identifying emerging hotspots, and predicting trends for disease. Aggregation of single predictions will help health officials gain actionable insights into pneumonia prevalence, which can be highly targeted in interventions and allocation of resources. Moreover, the model is automated, scalable, and deployable in high-demand settings toward a proactive approach to respiratory disease management. In summary, this article presents the intersection of CNN diagnostics and epidemiology, revealing the role of predictive analytics in public health preparedness and response.