Pneumonia is an infection which is caused by bacteria or viruses. Early diagnosis is critical to treat the disease successfully without delaying the treatment much. In most of the cases and as per the usual process the patient with pneumonia-like symptoms can be dragonized via frontal and lateral chest x-ray images, which are then seen over by the naked eye by doctors or radiologists. The diagnoses can be misleading and confusing as the appearance of the disease can be unclear in X-ray images and can put the doctor in a dilemma, as the features may not be visible clearly via naked eyes. That is why computer-aided diagnosis is generally required to guide clinicians. The model is based upon the convolutional neural network architecture, wherein pre-processed images are fed to the developed network layers and trained to provide us results with high accuracy of 94.3%, a precision rate of 93.18%, recall of 98.20% and an F1 score of 95.63%. The objective of the work is to design a model that can provide fast and accurate analysis which not only may save diagnosis cost, but also provide invaluable time for the doctors to begin the treatment if the disease is detected early.