Zero to your First Logic App AI Agent using Agent Loop in 30 Minutes!
Unlock the power of Azure Logic Apps Agent Loops with this detailed, step-by-step tutorial designed for professionals building AI-driven workflows in Azure. In this video, Stephen Thomas walks you through the brand-new Agent Loops feature (currently in public preview), demonstrating how to integrate AI agents that can utilize Microsoft connectors, third-party APIs, and HTTP calls—all inside Logic Apps.
What you’ll learn:
How to set up Azure OpenAI with GPT-4.1 for Logic Apps
Creating a Standard Edition Logic App and understanding cost considerations
Designing and planning a dynamic agent workflow (example: “High Card Wins” card game)
Configuring agent instructions, parameters, and autonomous triggers
Integrating connectors and HTTP tools, plus email notification setup
Testing, debugging, and best practices for request/response patterns
Referenced Links:
Blog Post with Agent Instructions - https://www.stephenwthomas.com/azure-...
Azure OpenAI Pricing - https://azure.microsoft.com/en-us/pri...
Azure OpenAI Model Deployment Guide - https://learn.microsoft.com/en-us/azu...
Key Highlights:
00:00 Introduction
01:14 Laying out the AI Agent Plan
04:22 Creating Azure OpenAI
08:35 Creating Chat GPT 4.1 Model as a Service
10:39 Create Standard Logic App and App Service Plan
13:04 Create Agent Loop Workflow (Public Preview)
13:43 First Look at Agent Loop Action
16:12 Setting up the Agent Loop
19:13 Adding 3 Separate Tools to Agent Loop
26:02 Adding Request Trigger
26:50 Add Response using Last Assistant Message
27:16 Fixing Design Time Errors
27:54 Testing the Agent Loop Logic App
29:24 Update Agent Instructions using AI
31:41 Next Steps
Remember to like and subscribe to @StephenWThomas for more tutorials on Azure Integration Services, Logic Apps, and Artificial Intelligence (AI)!
Check out my website and blog at https://www.StephenWThomas.com