Memgraph's MAGE 1.4 release enhances its open-source graph extension library by introducing support for graph machine learning, enabling users to classify nodes and predict new relationships using node classification and link prediction algorithms. The update includes a C++ API that simplifies memory management and enhances usability, making it more accessible to C++ developers. The integration of the igraph library and the k-means clustering algorithm further expands MAGE's capabilities, supporting a variety of graph algorithms and offering improved data handling through wrappers. The new features are designed to facilitate tasks such as service recommendation and fraud prediction, with support for popular frameworks like GraphSAGE, GAT, NetworkX, and cuGraph. Users are encouraged to update to the new version, contribute feedback on GitHub, and engage with the community via Discord for ongoing discussions and support.