Pavan Nutalapati

Title of the Talk :
Self-Healing Clouds: Agentic AI for Autonomous Security and Disaster Recovery at Planetary Scale

Abstract of Talk:

By 2026, cloud failures will no longer be isolated outages but AI-accelerated, multi-region cascade events that simultaneously impact security, availability, and trust. This keynote presents an AI-first, disaster-recovery-native cloud architecture where agentic systems continuously observe threat signals, predict failure propagation, and autonomously trigger policy-driven recovery across regions and services. Drawing from large-scale distributed cloud platforms, I introduce a model where AI agents unify threat intelligence, resilience scoring, and recovery orchestration, optimizing SLA, cost, and regulatory constraints in real time. The session explores how machine learning shifts disaster recovery from static runbooks to adaptive decision engines, enabling self-healing clouds that respond to cyber attacks, data corruption, and infrastructure instability without human intervention. This talk reframes cloud DR as an intelligent security control plane, positioning autonomous recovery as a foundational requirement for next-generation hyperscale systems.

Bio:
Pavan Nutalapati is a senior engineer with over 17+ years of experience designing and operating large-scale, mission-critical distributed systems. His work focuses on cloud security architecture, cross-region disaster recovery, and automation for hyperscale platforms, with hands-on experience building resilient systems that support millions of users. He has led and contributed to advanced initiatives spanning AI-assisted security, policy-driven recovery orchestration, and fault-tolerant cloud infrastructure, combining deep engineering rigor with applied research.

He currently works on AI-driven cloud resilience and disaster recovery automation, where he designs agent-based systems that predict failures, contain security incidents, and autonomously trigger recovery actions across regions. His work emphasizes optimizing availability, cost, and compliance in real time, helping enterprises transition from manual, runbook-based recovery to self-healing cloud platforms. Through this work, he contributes original approaches to making cloud infrastructure secure, adaptive, and resilient in the face of rapidly evolving cyber and operational threats.