google cloud Introducing major updates to maintain AIDevelopers of the Vertex AI platform for conceptualizing, designing, building, testing, deploying, and modifying AI agents in enterprise use cases.
New features announced today include additional governance tools for enterprises, enhanced ability to create agents with just a few lines of code, acceleration with a cutting-edge context management layer and one-click deployment, managed services for scaling production and evaluation, and support for agent identification.
agent builder, released last year is an annual Cloud Next event that provides a no-code platform for enterprises to create agents and connect them to orchestration frameworks such as LangChain.
Google agent development kit (ADK), which allows developers to build agents in “less than 100 lines of code,” and is also accessible through Agent Builder.
“These new features highlight our commitment to Agent Builder to simplify the agent development process and enable us to meet developers where they are, no matter which technology stack they choose,” said Mike Clark, director of product management for Vertex AI Agent Builder.
Build agents faster
One of Google’s pitches for Agent Builder’s new features is that it allows businesses to enhance orchestration while building agents.
“Building an agent from concept to working product requires complex orchestration,” Clark says.
New features shipped with the ADK include:
-
SOTA context management layers, including static, turn, user, and cache layers, give enterprises more control over agent context
-
Pre-built plugins with customizable logic. One of the new plugins allows the agent to recognize failed tool calls and “self-heal” by retrying the task with a different approach.
-
Additional language support in ADK (such as Go, Python, and Java launched with ADK)
-
One-click deployment through the ADK command line interface allows you to move agents from your local environment to live testing with a single command.
governance layer
Businesses require high precision. Safety; observability and auditability (what the program did and why). Maneuverability (control) of production-grade AI agents.
While Google included observability capabilities in local development environments at launch, developers now have access to these tools through a runtime dashboard managed by the agent engine.
The company said this will enable cloud-based production monitoring to track token consumption, error rates and delays. Within this observability dashboard, businesses can visualize the actions that agents take and reproduce issues.
The agent engine also adds a new evaluation layer that helps “simulate agent performance across a variety of user interactions and situations.”
This governance layer also includes:
-
Agent ID, Google says, “gives agents their own unique native ID within Google Cloud.”
-
Model Armor blocks prompt injections, screening tool calls, and agent responses.
-
Security Command Center. Administrators can build an inventory of agents to detect threats such as unauthorized access.
“These native identities provide a deep, built-in layer of control and a clear audit trail for all agent actions. These certificate-backed identities cannot be spoofed and are tied directly to the agent lifecycle, further enhancing security and eliminating the risk of dormant accounts,” said Clark.
agent builder battle
It makes sense for model providers to create a platform for building agents and deploying them into production. Competition depends on how quickly new tools and features are added.
Competes with Google’s Agent Builder OpenAIopen source agent development kitThis allows developers to create AI agents using non-OpenAI models.
Furthermore, recently, Announcing AgentKitfeatures Agent Builder, which allows businesses to easily integrate agents into their applications.
Microsoft has Azure AI Foundrylaunched around this time last year to create AI agents, AWS also offers Agent builder on foundation But Google hopes a series of new features will help give it a competitive edge.
But companies with proprietary models aren’t the only ones asking developers to build AI agents within their platforms. Enterprise service providers with agent libraries also want their clients to create agents on their systems.
Gaining the attention of developers and keeping them in the ecosystem with features that make it easy to build and manage agents is now a big battle among technology companies.
