The goal of this project is to build a neural network that takes an MNIST handwritten digit (0-9) image and a random number (digit 0-9) as inputs and returns the predicted class label (0-9) for the input image and its addition (sum) with the input random number as summed output (range 0-18) label as outputs.