What if troubleshooting a GKE workload was as simple as having a conversation?

Platform engineering has traditionally relied on commands, dashboards, YAML, runbooks, and deep Kubernetes expertise. But with AI entering the developer workflow, we're moving toward a new model where engineers can interact with infrastructure using natural language while still working within the guardrails of the platform.

In this practical, demo-driven session, we'll explore how Gemini can become a powerful copilot for platform engineers on Google Kubernetes Engine (GKE). Starting with a real-world broken workload, we'll move beyond traditional `kubectl`-based troubleshooting and progressively give AI the context it needs to understand code, Kubernetes configurations, and the live GKE environment.

Through a hands-on Codelab-style demonstration, you'll see Gemini CLI, the GKE Model Context Protocol (MCP), and AI-powered workflows in action from diagnosing failures and understanding cluster state to generating configurations and turning repetitive operational knowledge into reusable workflows.

You'll leave with a practical understanding of how AI can augment not replace platform engineers, and how teams can begin building smarter, more contextual, and developer-friendly cloud platforms on GKE.