Imagine generating stunning AI images from text prompts — all powered by Kubernetes, accelerated by GPUs, and designed with the help of Gemini.

In this talk and live demo, we’ll explore how to build, containerize, and deploy a full AI Image Generator using Stable Diffusion XL on Google Kubernetes Engine (GKE). You’ll see how Gemini assisted in creating the entire solution — from writing Python and Docker code to building the Streamlit UI and deployment YAMLs — demonstrating how modern AI tools can supercharge developer productivity.

We’ll then dive into the technical journey: provisioning GPU nodes, deploying workloads, scaling inference with Kubernetes, and pushing reproducible builds to Google Artifact Registry.

💡 Key Takeaways:

• Learn how to build and scale Stable Diffusion on Google Kubernetes Engine for real-time image generation

• See how Gemini accelerates AI app development, from code to deployment

• Understand GPU scheduling, autoscaling, and secure container management on GKE

• Watch a live demo: “A Kubestronaut riding a dragon in space” — rendered on GKE in seconds

• Whether you’re a cloud engineer, AI developer, or Kubernetes enthusiast, this session showcases how AI can help you build AI, turning ideas into deployable, cloud-native applications faster than ever.

• 🌐 Presented by KSUG.AI — the world’s most active Kubernetes + AI community.

+ food and networking 🍕

Presented by:

Yongkang He — the world’s first Kubestronaut — is the founder of KSUG.AI, home to one of the largest Cloud Native and AI user communities with over 200,000 members / followers worldwide.

He also created the Awstronaut and Kubestrong programs to unite and empower Cloud Native and Multi-Cloud professionals across the globe.

🔗 Follow him & Learn more at KSUG.AI