At the GKE Turns 10 Hackathon, I built the Sustainable Shopping Advisor, an AI-powered microservice that extends the Online Boutique with real-time sustainability intelligence. The idea started from a simple gap: e-commerce platforms rarely provide clear environmental insights, even though consumers are eager to make greener choices.
The solution was to create a containerized sustainability agent running on Google Kubernetes Engine. Instead of touching the original Online Boutique code, the agent connects through APIs using the Model Context Protocol (MCP). With the Agent Development Kit and Agent-to-Agent protocol, I was able to orchestrate specialized agents that analyze products and provide sustainability recommendations directly in the shopping experience.
The main challenge was enabling smooth interaction between the new microservice and the existing application without breaking anything. Deploying on GKE made it possible to containerize and scale the advisor, while Gemini models added intelligence to the analysis. In just a short time, I was able to demonstrate how an external AI “brain” can upgrade microservices with sustainability scoring and recommendations.
The Online Boutique now has an extra layer of AI-driven sustainability insights, showing how any e-commerce platform could adopt greener intelligence without complex rewrites. Moving forward, this foundation could evolve into richer sustainability metrics, improved recommender logic, and even a reusable plug-and-play service for other platforms.
Top comments (0)