AI assistants are everywhere—but most fall short when it comes to real developer workflows. They offer answers without context, and in the world of Kubernetes, that’s rarely enough. Lens Prism, the AI assistant embedded in Mirantis’ popular Lens IDE, is changing that.

Kyle Wheeler, General Manager for Lens at Mirantis, shares the thinking behind Prism: “Our strategy has been to essentially have a middle layer of pre-prompting that we do ourselves, to give context to the LLM.”

This pre-prompting layer feeds environment-specific data into the large language model (LLM), whether that’s OpenAI, Azure, or any OpenAI-compatible local model. That means the assistant knows what cluster you’re looking at, what namespace you’re working in, and even which pods are in play.

From Commands to Conversations

Rather than relying on developers to craft the perfect prompt, Lens Prism takes familiar developer questions like “What’s wrong with this pod?” and uses internal awareness to deliver accurate, context-rich answers.

“You can ask the same question your leadership asks—‘What’s wrong with this infrastructure?’—and actually get an answer,” says Wheeler.

This isn’t just a conversational upgrade; it’s a transformation in how developers debug and navigate their systems. Instead of trial and error in the terminal, Prism gives them guidance—rooted in real-time observability from Lens.

Personalized AI, At Scale

Beyond the default behavior, Prism can be customized to suit individual users or entire organizations. From simple settings like language preferences (“Speak only to me in German”) to deeply technical pre-prompts aligned with internal policies, the AI assistant becomes more than a chatbot. It becomes part of your organizational workflow.

“If your company has specific rules for how things are done,” Wheeler explains, “Lens Prism can be configured to abide by those rules.”

This bridges the gap between generic AI models and the unique needs of enterprise environments—especially those operating at scale across clusters.

A Tighter UI-AI Integration

While current functionality centers on diagnostics and conversational insights, the roadmap is ambitious. Wheeler shared plans to deeply integrate Prism’s suggestions with Lens’ user interface. In future releases, users will be able to click through from AI output directly into related logs, dashboards, and resource views.

As Wheeler puts it: “Lens Prism will surface information with links to that specific log or dashboard. You can go right there in the context of Lens.”

That means a developer can go from “What’s wrong?” to “Here’s the log” to “Now I’ve fixed it”—all without leaving the IDE.

AI That Understands Kubernetes

With its focus on visibility, observability, and direct action, Lens Prism offers a new standard for AI in developer tooling. It doesn’t replace terminal fluency—it complements it, especially for newer Kubernetes users or overworked platform engineers.

AI in dev environments must go beyond code autocomplete. It should reduce mental overhead and provide real-time, contextual help rooted in the systems developers are already navigating. With Prism, Lens is delivering just that.