Modern applications generate massive volumes of logs across microservices, APIs, identity systems, and cloud platforms but most organizations still struggle to convert this raw telemetry into meaningful security detections. Traditional rule-based alerting often misses subtle behavioral anomalies, while noisy signals overwhelm analysts and leave critical threats undetected.

This talk demonstrates how to build an AI Agent-enhanced XDR pipeline that transforms application logs into actionable defense signals. We will explore how machine learning, behavioral analytics, and LLM-based reasoning can uncover API abuse, authentication anomalies and other attacks that rarely surface in conventional AppSec testing.