Meet Liz: The Rancher AI Assistant

Liz is a context-aware AI agent integrated into the SUSE Rancher Prime environment. Unlike general-purpose AI, Liz is designed to understand the live state of your clusters, namespaces, and workloads.

Available as an extension, Liz resides within your workspace to provide real-time guidance where you need it. Whether you are investigating a failing deployment or trying to understand a new Kubernetes concept, Liz uses the Model Context Protocol (MCP) to bridge the gap between your operational data and actionable insights.

Whether you need a Site Reliability Engineering (SRE) opinion or the help of a Rancher expert, Liz can provide you insight to help you make the right decision, faster. The power of Liz lies in its integration with the Rancher UI and its ability to adapt to enterprise security requirements.

Key Components

  • AI Agent: The bridge between your LLM and the Rancher MCP. The AI agent component enforces guardrails to help protect your environment from LLM hallucinations. It also contains a built-in RAG (Retrieval-Augmented Generation) system to provide the LLM of your choice with the latest Rancher documentation and knowledge base.

  • Rancher MCP: The Rancher Model Context Protocol (MCP) is the component in charge of interacting with the Rancher API. It uses the Rancher public API mechanism and exposes it using tailored tools to provide the best user experience.

  • Rancher Assistant UI Extension: This component extends the Rancher UI for Liz. It goes beyond a simple chat window by embedding Liz where it makes sense in the UI. Integrating the assistant behind error messages and status badges for immediate, contextual support.

Key Features

Flexible Model Sovereignty

Recognizing that security is paramount, Liz offers freedom of choice regarding the underlying AI engine.

  • Connect to local or air-gapped LLMs for maximum data privacy using Ollama.

  • Use public engines for high-performance reasoning, such as:

    • OpenAI’s ChatGPT

    • Google’s Gemini

    • AWS Bedrock

Why an AI Assistant?

As Kubernetes environments scale, the "cognitive load" on operators increases exponentially. Liz addresses three critical challenges in modern platform operations:

  1. Closing the Skills Gap: Kubernetes has a notoriously steep learning curve. Liz provides immediate, expert-level support to engineers who may be unfamiliar with specific resources or error types, reducing the reliance on senior "hero" engineers.

  2. Accelerating Resolution Time: Instead of manually examining logs or searching external documentation, operators can ask Liz to summarize the state of a namespace or explain a specific failure, reducing the mean time to resolution (MTTR).

  3. Proactive System Awareness: Liz enables a proactive operational stance. By asking questions like "Which deployments are currently unhealthy?" or "Summarize the health of this cluster," teams can identify and fix issues before they escalate into outages.

While Liz is a powerful assistant, it’s designed to extend human knowledge, not replace it. Users should verify AI recommendations and follow their organization’s standard security practices when applying suggested fixes.