In the field of Data Science, traditional Machine Learning models can achieve excellent predictive performance. During the learning process the algorithms (at any level of complexity) adapt themselves to the training dataset, identifying patterns of correlation among the variables in the training dataset and leveraging helpful correlation between input features and output target. This optimization process, based on increasingly sophisticated techniques, enables high predictive accuracy across a wide range of problem types and real phenomena. However, while Machine Learning algorithms detect correlations within the input dataset quite well, they fall short of capturing true cause-and-effect relationships. This limitation reduces their applicability and interpretability in various business scenarios. Approaching a definition, designing and implementing a data-driven model in a real scenario we are often interested in detecting causal relationships between input features and target, also to justify the physical behavior of the Machine Learning model. Causal Inference addresses this gap by enabling the identification of the actual impact of input variables on the target, using several specific techniques. In this talk, we will explore the Causal Machine Learning techniques, a synthesis between Machine Learning algorithms trained on real-world data and modern Causal Inference methods, using the Python library CausalML. The CausalML library is one of the most comprehensive and well documented open-source solutions in the field, implementing state-of-the-art algorithms for estimating causal effects and allowing seamless interpretation of powerful Machine Learning models. This session will introduce the fundamental concepts of Causal Inference, its application to Machine Learning algorithms, the main features of the CausalML library, and a real-world use case. By the end of the session, participants will be equipped to apply these techniques effectively to their own Data Science projects.