Google Cloud Launches Enterprise AI with Agent Studio

Agent Studio, our low-code interface, lets developers and business users alike build and test agents using natural language. The Gemini Enterprise app brings AI to your everyday work. We're bringing the power of the agents directly to your workforce through the Gemini Enterprise app.

To ensure everyone can tap into this technology, our no-code Agent Designer lets anyone build custom, trigger-based workflows without writing a single line of code. For more complex, multi-step business processes, new long-running agents can work autonomously in the background within secure cloud sandboxes while you focus on other things. With all these agents helping out, we've built a central, intuitive Agent Inbox so you can easily monitor, guide and manage exactly what your AI helpers are up to.

Our new eighth-generation TPUs are designed to power the AI era. Running millions of AI agents takes some serious computing muscle. Our AI Hypercomputer is a purpose-built system designed specifically for the massive scale of this new era, including the eighth generation of our custom TPU (Tensor Processing Unit) chips.

The TPU 8t is built to train AI models incredibly fast, while the TPU 8i is optimized for inference (actually serving up the models), delivering 80% better performance per dollar. We will also be among the first to offer the new NVIDIA Vera Rubin NVL72 systems, joining our existing lineup of NVIDIA GPUs and our super-efficient Google Cloud Axion processors. To take advantage of this powerful compute, you need to move data at lightning speed.

We unveiled the Virgo Network, a custom-built system to connect massive supercomputers, alongside storage breakthroughs like Managed Lustre, which can now move an incredible 10 terabytes of data per second. The new Agentic Data Cloud helps make sense of your data.