hands-on lab

Creating Agentic Reasoning & Action with Agent Designer on Google Cloud

Difficulty: Beginner
Duration: Up to 1 hour and 30 minutes
Students: 2
On average, students complete this lab in2h 25m
Get guided in a real environmentPractice with a step-by-step scenario in a real, provisioned environment.
Learn and validateUse validations to check your solutions every step of the way.
See resultsTrack your knowledge and monitor your progress.

Description

Agentic AI reasons toward a goal and decides when to use tools available. It also may ground answers in external evidence rather than guessing. You can apply agentic patterns with or without a formal “agent” construct.

In this hands-on lab, you’ll use Vertex AI Studio to configure a Vertex AI Agent to perform Reasoning and Action in Agent Builder. The agent will ask follow-up questions to assist in reasoning and, using a set of tools, construct a query to fulfill your request. Sub-Agents will be used to run the tools.

Learning objectives

Upon completion of this lab, you will be able to:

  • Understand how agentic AI differs from a traditional chatbot by including the roles of reasoning and action in intelligent systems.

  • Navigate the Vertex AI ecosystem (Vertex AI Studio, Agent Builder) and describe how these components integrate within Google Cloud.

  • Create and configure a first agent using Vertex AI Agent Builder (Agent Designer), including defining agent identity, instructions, model and basic tool/grounding behavior.

  • Apply best practices for designing and scoping a first agent (clear purpose, bounded behavior, observable outcomes) and understand how templates/connectors accelerate future development.

Intended audience

  • Cloud Solution Architects

  • Platform / DevOps Engineers exploring AI capabilities on GCP

  • Developers and Technical Consultants new to Vertex AI

  • Innovation, R&D, and AI Enablement teams evaluating agentic approaches

  • Technical Program Managers supporting AI platform adoption

Prerequisites

Basic familiarity with the Google Cloud Console (projects, navigation, permissions)

General understanding of what an LLM is (no ML training required)

The following content can be used to fulfill the prerequisites:

Environment before

Environment after

Covered topics

Hands-on Lab UUID

Lab steps

0 of 3 steps completed.Use arrow keys to navigate between steps. Press Enter to go to a step if available.
  1. Signing In to the Google Cloud Console
  2. Creating Agent Reason-Act workflow in Designer
  3. Testing the Agentic Workflow for Reasoning Action