This repository provides the content and sample code for a 75-minute workshop focused on building a code-first conversational agent using Azure AI Foundry Agent Service. It targets developers and data professionals looking to leverage Azure's AI capabilities to analyze business data and create intelligent applications. The workshop enables participants to build, explore, and guide an agent for tasks like sales data analysis.

How It Works

The workshop guides users through creating an agent application with Azure AI Foundry Agent Service. It emphasizes the use of "tools" within the service and effective "instructions" to direct the Large Language Model (LLM) for specific tasks, such as querying sales data. This approach allows for the creation of specialized, data-aware conversational agents without extensive custom LLM fine-tuning.

Quick Start & Requirements

• Primary install/run command: The workshop is designed to be run within GitHub Codespaces.
• Prerequisites: An Azure subscription (free tier sufficient for workshop), a GitHub account, and sufficient Azure credits/quota (estimated < USD$1.00 for a typical run).
• Setup: Designed for quick setup within the Codespaces environment.
• Links: Workshop Guide: Contoso Sales AI Agent

Highlighted Details

• Focuses on a "code-first" agent development approach.
• Demonstrates practical application of Azure AI Foundry Agent Service for business data analysis.
• Includes sample application code that can be modified or replaced.

Maintenance & Community

• Contributions are welcomed via GitHub issues and pull requests, subject to a Contributor License Agreement (CLA).
• Follows the Microsoft Open Source Code of Conduct.

Licensing & Compatibility

• The repository content and sample code are provided under a permissive license, allowing for modification and use. Specific license details are not explicitly stated in the README but are typical for Microsoft open-source projects.
• The code is intended to showcase Azure services and is not recommended for production without additional security features.

Limitations & Caveats

Some features used in the repository are in preview, meaning they are provided without a service level agreement and may not be suitable for production workloads due to potential instability or limited capabilities.