Pondurance Launches Agentic AI SOC for Autonomous MDR

Pondurance has announced the general availability of Kanati, an Agentic AI platform designed to power its managed detection and response (MDR) services and redefine how security operations centers (SOCs) function. The new system introduces an AI-native operating model where automation handles high-confidence threats in real time, while human analysts shift into supervisory and strategic roles. As cyberattacks increasingly operate at machine speed often leveraging AI traditional SOC models struggle to keep pace due to manual triage processes and alert fatigue. Pondurance’s Kanati aims to address this gap by enabling faster detection, investigation, and response across enterprise environments.

According to the company, Kanati processes massive volumes of security telemetry, correlating endpoint, network, cloud, and identity data to identify threats with greater speed and accuracy. By autonomously handling high-confidence incidents, the platform significantly reduces response times while minimizing the burden on human analysts. This allows security teams to focus on more complex or ambiguous threats that require deeper expertise. Initial performance benchmarks highlight the impact of this approach. Kanati is reported to accelerate threat analysis by up to 90%, reduce false positives by 80%, and investigate alerts in under two minutes on average. Additionally, it enhances contextual threat analysis and correlation, enabling faster identification of vulnerabilities before they can be exploited.

Unlike traditional SOC workflows that rely heavily on manual processes, Kanati uses a coordinated system of AI agents that operate continuously across the entire threat lifecycle. These agents ingest and normalize data in real time, conduct multi-step investigations, and execute verified containment actions such as isolating endpoints or restricting compromised identities. At the same time, the system generates detailed, audit-ready reports for every incident.