← Research
● PublishedIEEE SCEECS 2026 · Bhopal2026

Agentic AI Framework for Autonomous and Self-Managing Cloud Services

V. Shukla, V. Shukla, A. Soni, Atul, M. K. Das, R. Tiwari

Agentic AI Framework for Autonomous and Self-Managing Cloud Services

Rule-based Kubernetes autoscaling struggles when workloads shift unpredictably. This framework uses multi-agent reinforcement learning so autonomous agents orchestrate cloud-native apps together — beating traditional orchestration on resource usage, latency and resilience in simulation.

Static, rule-based Kubernetes autoscaling breaks down when workloads move unpredictably. Our framework uses multi-agent reinforcement learning (MARL) so agents orchestrate cloud-native apps together — and in a simulated cluster it beats conventional orchestration on resource use, service latency and operational resilience. Full details via the DOI above.