See full agenda Tue Oct 6 / 10:40 - 11:10 CEST

The Cybersecurity World Model: From Noisy XDR Incidents to an AI-Native SOC

Modern XDR platforms generate large volumes of incidents, making it difficult for SOC analysts to identify what truly matters. This presentation introduces the Cybersecurity World Model - a foundational AI model trained specifically on cybersecurity data to understand incidents, their context, and their relationships. We will discuss the journey from traditional detection approaches to cybersecurity-native foundation models, the infrastructure and data that made this possible, and how the resulting model delivers strong performance with a compact architecture. Finally, we will demonstrate how this technology enables AI-native and agentic SOC workflows, bringing the industry closer to effortless cybersecurity.

AI Tech

More presentations from Filip Mazan: Q&A

More presentations from Jan Macek: Q&A

07. Filip Mazan - Sr. Manager of Advanced Threat Detection and AI Research
Filip Mazan Sr. Manager of Advanced Threat Detection and AI Research
08. Jan Macek - Senior Software Engineer
Jan Macek Senior Software Engineer
See full agenda