Uses a convolutional neural network to detect pneumonia on x-ray images of chests. We are able to classify correctly 96% of the time. Our most important success criteria was a low false negative rate, in which our model managed to stay below 5%.