Segmentation tasks can be very easy to humans but very tough to machines. For us is quite simple to identify what objects are inside an image and their edges, however a machine doesn't know the meaning of pixels of an image, only their position and intensity. Recent advances in AI are bringing new capabilities to those models and make them learn from data, however to achieve good results is required to follow some steps as well as choose the correct tool for each type of problem. The aim of this talk is to introduce to the audience which categories AI is organized and which one fits into the segmentation task, which deep learning structures are commonly used and the steps to create an image segmentation model.