Machine learning algorithms excel at discovering statistical dependency patterns in data. Those can be exploited to produce extremely accurate predictions, often beyond human ability. Nonetheless, for decision-making tasks that inform action policies, all statistical dependencies are not equivalent. Some are spurious and uninformative, while others stem from causal relationships and are actually informative. In this talk, we will see how we can combine knowledge of the causal structure of a data-generating distribution with machine learning to accurately estimate the effect of actions. We will also discuss some key problems, where there is a high potential for synergy between the fields of machine learning and causal inference and discuss some recent work. The talk will be followed by a practical session, where participants will get to experiment with the presented methods.