The JNUC session catalog is subject to change. Please remember to check your schedule once on-site.

In an era where phishing attacks are becoming increasingly sophisticated, traditional detection methods often fall short of identifying complex threats. In our mission to protect customers against evolving cyber threats, our approach combines the strengths of three specialized groups: security analysts, developers and data scientists. Our security analysts manage and monitor billions of requests and extensive blocklists, working around the clock to detect, investigate, and respond to potential threats. Supporting this frontline defense, our developers have built a powerful analytical pipeline that enriches raw data—transforming it into actionable insights that significantly enhance threat hunting capabilities. Meanwhile, our data scientists design, train and continuously improve machine learning models to boost detection accuracy and proactively identify emerging threat patterns. Our cross-functional collaboration focuses on delivering intelligent, scalable and proactive cybersecurity solutions to keep our customers safe. This technical session showcases how this unified approach has revolutionized our phishing detection capabilities through advanced HTML content analysis and machine learning and explores an innovative approach to enhancing phishing detection. We'll demonstrate the architecture and methodology of extracting meaningful features and patterns from HTML content to improve threat detection accuracy while maintaining performance. Attendees will learn how our approach combines human expertise with machine learning capabilities to create a more robust and intelligent security solution. We'll share practical insights into building and maintaining such a system, including technical challenges overcome and lessons learned in combining traditional security approaches with modern machine learning techniques.