Descriptions:
Google Cloud developer advocate Nardini walks through a complete end-to-end demonstration of building and deploying an application using Claude Code on Google Cloud, structured around five team personas: product manager, UX/UI designer, software engineer, security engineer, and data analyst. Each persona maps to a distinct phase of the software development lifecycle, with Claude Code augmenting every stage from ideation through production deployment.
The demo builds a live feedback application and deploys it using a stack of Google Cloud services: Cloud Run for serverless hosting, Firestore for real-time response storage, and BigQuery for analytical processing and dashboard integration. Two recently released Google Cloud integrations take center stage — the Developer Knowledge API with its associated MCP server, which gives Claude Code direct access to up-to-date GCP documentation to determine optimal architectures, and Google Cloud Skills for automated task execution. Setup uses Application Default Credentials (ADC), eliminating the need to manage or rotate API keys manually, and a new Claude Code wizard automates project and region detection.
The talk makes a practical case for running Claude models on Google Cloud specifically: per-token billing with no message caps, provisioned throughput for production workloads, and regional model availability. Nardini positions the integration as lowering the barrier for enterprise teams to deploy AI-assisted development workflows without needing deep Google Cloud expertise — letting Claude Code handle architecture decisions through real-time documentation access.
📺 Source: Claude · Published May 08, 2026
🏷️ Format: Tutorial Demo
