On April 22, 2026, Google Cloud introduced the Gemini Enterprise Agent Platform — a new comprehensive platform designed for creating, scaling, managing, and optimizing AI agents. This initiative is intended to significantly simplify the development of reliable and secure generative AI-powered business tools, the creation of which previously required significant engineering effort and a high tolerance for errors. The platform represents an evolution of Vertex AI, combining existing model selection and development capabilities with new features such as agent integration, DevOps practices support, orchestration tools, and enhanced security measures. The Gemini Enterprise Agent Platform is positioned as a single hub for technical teams, enabling them to transform companies' products, services, and operations.
Platform users gain access to over 200 leading models via Model Garden. These include Google's latest developments, such as Gemini 3.1 Pro, Gemini 3.1 Flash Image, and Lyria 3, as well as open models like Gemma 4. The platform also supports third-party models, including Claude Opus, Sonnet, and Haiku from Anthropic, providing flexibility in choosing the optimal tool for a specific task. All future Vertex AI services and updates will now be provided exclusively through the Agent Platform, emphasizing its central role in Google Cloud's AI agent development strategy.
This transition aims to ensure the managed, secure, and scalable deployment of autonomous agents in an enterprise environment, capable of operating with the independence and reliability of a team member. The platform offers a variety of tools for developers: from the low-code visual interface Agent Studio to the code-centric Agent Development Kit (ADK). The Agent Runtime component provides support for long-running, stateful agents for several days thanks to Memory Bank. For management and control, Agent Identity, Agent Registry, and Agent Gateway are provided, and for optimization — Agent Simulation, Agent Evaluation, and Agent Observability. For example, Burns & McDonnell is already using the Agent Platform, specifically ADK, to transform corporate knowledge.
They are creating an AI agent that transforms decades of project data into real-time actionable insights, demonstrating the platform's potential for responsibly scaling innovation through a combination of deterministic business rules. Technical teams can start working with the Gemini Enterprise Agent Platform now by visiting the Google Cloud console. This will allow them to explore new capabilities and begin creating production agents, integrated through the Gemini Enterprise application. Such integration ensures a close connection with IT operations, guaranteeing the necessary control, management, and security when scaling AI solutions.
Sources
Replies (0)
No replies in this topic yet.
