This research introduces a novel method for automated pulmonary diseases detection and risk assessment using convolutional neural network (CNN) models hosted in the cloud. Pulmonary diseases are on the rise, and with them come the need for better diagnostic tools. Combining state-of-the-art machine learning methods with cloud computing infrastructure might lead to accessible and scalable healthcare solutions. The proposed CNN architecture is designed to efficiently extract features from chest radiographs, allowing for more precise illness classification and risk stratification. Using open-source data and cloud computing, the model is trained and tested on a wide variety of lung diseases and disorders, such as chronic obstructive pulmonary disease (COPD), pneumonia, and TB. Experiments show that the technique outperforms conventional diagnostics, with excellent sensitivity and accuracy across a variety of illness types. Optimizing diagnostic efficiency and enabling rapid treatments are made possible with cloud-based deployment, which also allows for real-time analysis and easy connection with current healthcare systems. This method might help with healthcare spending, patient outcomes, and the ability to provide healthcare remotely in areas with limited resources. It highlights might revolutionize automated lung disease detection and risk assessment, which in turn could improve customized medication and the management of population health. © 2025 Elsevier B.V., All rights reserved.