Solo.io Launches Kagent Enterprise for AI-Ready Kubernetes

Kagent enterprise makes Kubernetes AI-ready with context-aware connectivity and runtime.

Solo.Inference Orchestrator (IO), a provider of cloud native connectivity and infrastructure, announced the launch of Kagent enterprise on September 15, 2025, in Cambridge, Massachusetts. This platform aims to support Artificial Intelligence (AI) and agentic applications deployed on Kubernetes by addressing security, observability, resiliency, and IT governance challenges that often hinder AI projects from progressing beyond pilot stages.

Kubernetes serves as a robust base for cloud native applications, yet it lacks the contextual awareness necessary for managing Generative AI (GenAI) and agentic applications. Kagent enterprise addresses this limitation by adding context-aware networking and runtime components, which support agent-native protocols, including Model Context Protocol (MCP) and Agent-to-Agent (A2A) communication. This enhancement facilitates the implementation of AgentOps across various agent frameworks, such as the Agent Development Kit and Langchain.

Idit Levine, CEO and Founder of Solo.IO, emphasized the challenges organizations face in transitioning AI agents from development to production. “Kagent enterprise fills these important gaps by effectively and securely transforming cloud native infrastructure into agentic infrastructure,” Levine stated.

Kagent enterprise provides context awareness at multiple levels of the agentic infrastructure, which includes:

• Context-aware networking: This feature includes agentgateway, a data plane tailored for agentic AI connectivity, supporting MCP, A2A, and protocols from leading Large Language Model (LLM) providers, developed by Solo.IO and contributed to the Linux Foundation.
• Context-aware runtime: Kagent enterprise adds a runtime layer that allows Kubernetes to process workloads in a context-aware manner, requiring advanced policies for agent identity, as well as improved failure handling and observability for monitoring agent-tool interactions.
• Context-aware platform: The platform integrates connectivity and runtime features with a management plane that enables teams to oversee agentic infrastructure through an AgentOps dashboard, offering visibility of interactions between users, agents, tools, and LLMs.

Paul Nicholson, Research VP at IDC, remarked on the need for enterprises to develop a solid foundation for AI workloads to support scalability and security. “It is essential these applications have a foundation for security, observability, and scalability,” he stated, noting growing enterprise activity in deploying AI-enabled applications in the upcoming year.

Kagent enterprise aims to support platform teams in leveraging agentic AI more effectively, ensuring quality and security in production environments.