Pneumonia is a serious respiratory disease and it still remains one of the causes of the highest rates of morbidity and mortality worldwide. Effective treatment would require early and accurate diagnosis. This paper introduces a new method, which uses Large Language Models (LLMs) and Google Cloud Platform(GCP) and, in this case, Vertex AI to identify pneumonia in medical imaging data. The main goal of our project is to balance the high precision of the deep learning methods and automation and scalability of cloud computing to develop a fast and effective system of identifying pneumonia. The suggested strategy will be to train and deploy deep learning models on X-ray data of the chest by using Vertex AI Pipelines. Some of the automation steps involved in this framework include data preprocessing, model training, evaluation and deployment. With the GCP services of Cloud storage, big query and the AI platform, we are able to create an all cloud-native, scalable, and cost-effective solution that can be used both in clinical and remote healthcare applications. Our experimental evidence indicates a high level of accuracy in classification, consistent inference in real time and high performance. These results suggest that the combination of LLMs and cloud-based AI has a high potential to enhance medical imaging diagnostics and provide intelligent and scalable and deployable healthcare solutions.