The COVID-19 pandemic and current global health challenges such as pneumonia have reinforced the urgent need for accurate and rapid diagnostic solutions. Chest X-ray is a readily availablemodality for diagnosis of respiratory diseases, although its manual interpretation may be time-consuming and subjective. In this paper, we present a deep learning-based multiclass classification framework to categorize chest X-ray images into COVID-19, pneumonia, and normal. The dataset is composed of labeledX-ray images for the three classes preprocessed and augmented for an improved generalize model. A Convolutional Neural Network (CNN) model is trained to explicitly extract distinctive radiological patterns, which leads to high classification performing and excellent generalization ability to unseensamples. Evaluation metrics likeaccuracy, precision, recall and F1-score are employed for measuring model performance. The trained model are also deployed as an interactive web-app - that provides the ability for real-time diagnostic predictions via a humanfriendly user interface. This study proves the efficacy of deep learning in automated radiological diagnostics and provides an accessible system for clinical decision-support system, particularly in low-resourcehealth-care settings.