Skip to lesson

Google CloudProfessional Agentic Architect2026-091.1 Configuring agentic workflows and behavior using low-code tools

Choose a low-code agent builder

Completed by the knowledge check

Learning objective

Choose between Gemini Enterprise Agent Designer and CX Agent Studio from workflow and integration requirements. Google Cloud documentation now redirects the Agent Designer page to Workflow Builder; this lesson uses the exam term and notes the current product name.

Core idea

Choose the product from the operating environment:

  • Use Gemini Enterprise Workflow Builder (formerly Agent Designer) for employee-facing agents and workflows delivered in the Gemini Enterprise web app, especially when business users need no-code or low-code automation across enterprise data and tools.
  • Use CX Agent Studio for customer-experience agent applications that need conversational channels, hierarchical agents, customer-service integrations, or reuse of deterministic Dialogflow CX flows.

Both products can build agentic behavior. The deciding factor is not “which has AI,” but who interacts with the agent, where it runs, which systems it must connect to, and how much conversation and channel control the workload requires.

How it works

Google Cloud describes Workflow Builder as a no-code and low-code platform inside Gemini Enterprise. It creates chat agents and workflows through natural-language prompts or a visual editor. It can connect to governed data sources and tools, simulate workflows, publish them to organizational users or groups, and schedule recurring execution.

Within Workflow Builder, distinguish its two outcomes. A chat agent supports open-ended, user-driven conversation. A workflow coordinates a defined series of steps and can mix deterministic logic, model-based work, and human approval. This makes the product suitable for internal tasks such as employee onboarding, research assistance, or scheduled report preparation.

CX Agent Studio organizes a customer-experience application as a root steering agent and specialized subagents. Its current Agents documentation describes tools, deployment channels, sessions, escalation, multilingual interaction, and models tuned for customer experience. Evaluate it when the architecture needs voice or chat channels, service handoff, customer-session analytics, or contact-center integration.

CX Agent Studio can also import a Dialogflow CX flow as a flow-based agent. The flow integration hands control to the flow until END_SESSION and maps variables to and from flow session parameters. Google recommends flows for highly deterministic business logic such as sequential collection, validation, and authentication based on session parameters. This path matters when an organization already owns tested CX flows or needs explicit conversational state machines.

Before choosing, list the required surface, identities, connectors, channels, deterministic controls, deployment boundary, telemetry, and handoff behavior. Validate regional availability, quotas, and connector support in current product documentation rather than assuming feature parity.

Key distinctions and trade-offs

Employee workflow versus customer conversation: Workflow Builder is embedded in the Gemini Enterprise experience and governed by the tools and connectors an administrator provisions. CX Agent Studio targets customer-experience agent applications and their deployment channels.

Visual workflow versus CX flow: A Workflow Builder workflow can combine AI steps and defined actions. A Dialogflow CX flow is a state-machine structure with pages, parameters, routes, and event handlers. Similar diagrams do not imply the same runtime semantics.

New build versus flow reuse: CX Agent Studio can integrate existing Dialogflow CX flows, but Google warns that it is not a drop-in replacement for tightly coupled playbooks and flows. Reuse reduces migration risk while retaining session-parameter coupling that must be managed.

A general internal platform can speed broad employee automation. A specialized CX platform provides deeper customer-channel behavior but brings additional conversation, deployment, and integration design. Do not choose solely from a single demo or UI feature.

Architecture / decision scenario

A retailer has two projects. The first lets merchandising employees summarize supplier emails, query approved Drive content, and schedule a weekly catalog report. The team chooses Workflow Builder because the users work in Gemini Enterprise and administrators already govern those connectors.

The second project handles customer calls about deliveries, must preserve session context, escalate to agents, and reuse an existing identity-verification flow. The team chooses CX Agent Studio with the CX flow imported as a flow-based agent.

Using one product for both would hide different requirements. The shared need for Gemini reasoning is less important than the delivery surface and integration contract.

Common exam traps

  • Treating Agent Designer and Workflow Builder as separate current products when the official Agent Designer URL redirects to Workflow Builder.
  • Choosing CX Agent Studio merely because the workload contains conversation.
  • Choosing Workflow Builder for a customer voice workload without validating its required channels and contact-center integrations.
  • Assuming a CX flow and a model-driven agent have identical control behavior.
  • Ignoring existing Dialogflow CX assets and the migration cost of tightly coupled flows.
  • Assuming both products support the same connectors, regions, quotas, or deployment paths.

What to remember

  • Map Agent Designer in the exam to the current Workflow Builder documentation.
  • Choose Workflow Builder for governed Gemini Enterprise chat agents and business workflows.
  • Choose CX Agent Studio for customer-experience channels, sessions, escalation, and CX-specific integration.
  • Reuse Dialogflow CX flows for deterministic logic when their lifecycle fits.
  • Verify current product capabilities instead of assuming feature parity.

Knowledge check

References