In this workshop we explore the journey from relational data to actionable intelligence. Today’s security teams face a daunting "needle in a haystack" challenge, drowning in millions of isolated alerts and logs from traditional tools that miss the connected narrative of modern, sophisticated attacks. The core limitation lies in "flat," tabular data, which treats entities like users, IPs, and files as separate rows, stripping away the critical relationships that tell the true story of a threat. This is why relationships must be a first-class citizen in cybersecurity. By modeling an entire digital environment as a graph—with nodes representing users or devices and edges capturing their connections—we can finally see the hidden network. However, to decode complex patterns within this web, we need an intelligent analytical layer: Graph Machine Learning (GML). GML is a class of techniques designed to learn from graph structures, where algorithms understand a node’s context and role by analyzing its neighborhood, much like assessing a person’s influence not just by their title, but by their connections and community.
