MLJ is the preeminent all-in-one machine learning framework for Julia. In many ways, you can think of it like Python’s scikit-learn. In fact, MLJ provides a wrapper around scikit-learn’s models via PyCall. Thanks to Julia’s intelligent language design, MLJ makes it easy to add your own machine learning algorithms by implementing its simple interface. MLJ is thus able to provide a single unifying API for building, training, and evaluating a large selection of models taken from a variety of packages. Here, we will demonstrate its use on the famous Iris dataset.