Agentic AI for Autonomous Performance Engineering: An MCP-Driven Framework for Continuous Optimization in Multi-Tenant Cloud Systems

Authors: Pradeep Kumar

DOI: https://doi.org/10.37082/IJIRMPS.v13.i5.232766

Short DOI: https://doi.org/g98pk4

Country: US

Full-text Research PDF File:   View   |   Download

Abstract: This paper presents an Agentic AI and Model Context Protocol (MCP)-driven framework for autonomous performance engineering, shifting paradigms to self-adaptive, AI-driven optimization integrated with AIOps principles. By leveraging MCP an open protocol for secure context sharing, tool invocation, and modular interoperability the framework automates the full performance engineering lifecycle: from telemetry ingestion and anomaly-aware test generation to execution, root-cause analysis, and iterative remediation. A collaborative ensemble of five specialized Agentic AI agents Discovery, Authoring, Runner, Analyst and Publisher is orchestrated via MCP to enable goal-oriented reasoning and adaptive actions.

Keywords: Agentic AI, Performance engineering, AIOps, Autonomous testing, MCP framework, Continuous optimization, multi-tenant cloud

Paper Id: 232766

Published On: 2025-10-31

Published In: Volume 13, Issue 5, September-October 2025

Share this