Workflow Builder overview

Workflow Builder is an interactive no-code, low-code platform for creating, managing, and turning on agents in your Gemini Enterprise web app.

With Workflow Builder, you can:

  • Create agents using natural language prompts or the interactive visual editor.
  • Build and orchestrate deterministic or non-deterministic agent workflows using triggers, steps, and actions.
  • Connect your agents to various data sources (including Google, third-party, and custom data sources) and tools.
  • Preview and simulate agent behavior in the integrated testing environment.
  • Publish or turn on, and share agents with specific users or groups within your organization.
  • Schedule agent executions to run tasks on a recurring basis.

Agent types

The following table lists the type of agents that you can create using the Workflow Builder:

Agent type Description Best for
Chat agents Designed for conversational interactions to complete specific tasks. Responses depend on user input and are primarily non-deterministic. Can be single-step or multi-step agents.

For more information about chat agents, see Chat agent overview.

  • Open-ended conversations without fixed steps. For example, a brainstorming assistant for project ideas.
  • Dynamic tasks where user input drives the flow. For example, a customer support agent where the next question depends on the user's issue.
  • Q&A, data lookup, or personalized assistance. For example, checking Jira ticket status or summarizing a specific email thread.
Workflows Executed as a sequence of steps or actions, combining AI automation with human intervention. Logic can be deterministic or non-deterministic.

For more information about workflows, see Workflow overview.

  • Multi-step processes with defined sequence logic. For example, automating an employee onboarding checklist across tools like Gmail and Google Calendar.
  • Tasks with AI automation and human approvals. For example, drafting a report with AI that requires a user sign-off before publishing.
  • Complex orchestrations with conditional logic. For example, routing sales leads to different teams based on region and budget.

Agent controls and security

Gemini Enterprise provides centralized oversight and granular security for all AI agents within an organization. While Workflow Builder empowers employees and non-admin users to build and deploy custom agents to automate their specific workflows, these actions remain strictly governed by administrative policies. Specifically, the actions and tools available to any agent are limited to the data connectors, tools, and permissions provisioned by the Gemini Enterprise administrator. This ensures that agents operated by end-users can only perform tasks and access information that have been explicitly authorized, maintaining enterprise-grade compliance and data integrity.

What's next