In this work, neural networks and deep learning were used to equalize and decode multiple input multiple output (MIMO) systems. The initial focus was on learning equalization using neural networks with a fixed decoder. Subsequently, the learned equalizer was fixed and an NN was trained to function as a decoder.