Google Renamed Vertex AI. I Was There When the Original Strategy Was Shaped.

Google renamed Vertex AI last week. They're calling it the "Gemini Enterprise Agent Platform."

I was a founding member of Google Cloud AI. I spent 9 years building the enterprise AI infrastructure that this rebrand sits on top of.

Three things this rename tells you:

First — Google is betting that owning the full stack from chip to inbox wins. TPU 8th gen for training, Gemini for reasoning, Agent Platform for orchestration, Workspace for distribution. Kurian said it explicitly: "Other vendors hand you the pieces, not the platform."

Second — the A2A protocol reaching 150 organizations in production is more important than the rebrand. Microsoft, AWS, Salesforce, SAP, and ServiceNow are routing real agent tasks through A2A. Google positioned A2A as the layer above MCP: MCP connects agents to tools, A2A connects agents to each other across organizational boundaries.

Third — 75% of Google Cloud customers are now using AI products. 330 customers each processed over 1 trillion tokens in the past 12 months. 16 billion tokens per minute. Those aren't experimental numbers.

What Kurian didn't say: Google's agent platform now hosts 200+ models including Anthropic's Claude. They're conceding that model exclusivity is over. The platform play only works if enterprises can bring any model.

The risk: complexity. The keynote announced Agent Studio, Agent Designer, Agent Registry, Agent Identity, Agent Gateway, Agent Observability, Workspace Studio, and Agent Garden — in a single keynote. Brilliant engineering, ambitious scope, and a naming taxonomy that makes enterprise buyers reach for Advil.