Google Cloud’s Vertex AI Agent Builder, which debuted at Cloud Next last year, has just received a major refresh that is aimed at keeping developers inside the Google ecosystem for end‑to‑end agent development. The update adds a new observability dashboard, one‑click deployment, and a suite of governance tools that let enterprises scale production workloads while retaining tight control over agent behavior. At its core, the Agent Development Kit (ADK) remains the workhorse, enabling developers to prototype agents in under 100 lines of code in languages such as Python, Java, and the newly supported Go. The kit now ships with state‑of‑the‑art context‑management layers—Static, Turn, User, and Cache—that give teams granular control over what information the agent can see at any moment, and a library of pre‑built plugins that can, for example, detect failed tool calls and automatically retry with a different strategy.
Observability and governance have moved from a local‑only feature set to a fully managed runtime dashboard called Agent Engine. From the cloud console, users can monitor token consumption, error rates, and latency, and drill into the sequence of actions an agent takes to reproduce failures. The new Evaluation Layer lets teams simulate a wide range of user interactions before a model goes live. Security is bolstered by Agent Identities, which grant each agent a unique, certificate‑backed identity that can’t be impersonated, and Model Armor, which blocks prompt injections and screens tool calls. The Security Command Center provides a central inventory of agents, helping administrators detect unauthorized access and other threats.
With these enhancements, Google Cloud is sharpening its bid to outpace competitors such as OpenAI’s open‑source Agent Development Kit, Microsoft’s Azure AI Foundry, and AWS Bedrock. The real differentiator for the next‑generation agent builders will be how quickly they introduce developer‑friendly features and governance controls that meet enterprise compliance needs. By bundling rapid prototyping, one‑click deployment, and robust observability into a single platform, Vertex AI Agent Builder positions itself as a compelling choice for organizations that want to keep their AI agent development on the Google Cloud stack.
Want the full story?
Read on VentureBeat →