Step beyond basic chatbots and unlock the real power of AI with hands-on agent building. Learn to design, code, and deploy enterprise-grade AI agents using LLMs, APIs, and Microsoft Autogen. Walk away with skills that actually matter — and projects you can brag about.

• Build AI agents from scratch: Design, code, and deploy LLM-powered agents that actually do things, not just chat. .
• Engineer with Microsoft Autogen: Orchestrate multi-agent workflows and supercharge automation.
• Integrate real-world APIs: Connect your agents to live data, enterprise systems, and business apps
• Master prompt design & RAG: Craft effective prompts, use Retrieval-Augmented Generation, and make your AI use your data.
• Deploy in Azure like a pro: Set up, monitor, and scale agent solutions using Microsoft’s cloud tools.

Azure Subscription with access to OpenAI, AI Search, and Functions

Azure AI Studio or Foundry Portal access

Visual Studio Code (latest version)

REST Client / Postman / Jupyter for API testing

• Generative AI, LLMs, and their role in enterprise systems
• Intelligent agents: types and enterprise relevance
• Reactive vs tool-using agents
• LLM vs traditional ML: what’s different
• Azure AI Foundry and OpenAI service overview
• Use cases in Microsoft Teams, Outlook, and business operations

• Prompt design and task structuring
• Context windows, token management, and chaining outputs
• Using embeddings for personalization and memory
• RAG (Retrieval-Augmented Generation): workflows and architecture
• Azure AI Search + LLM: Using your own data
• LLMOps basics: deploying and maintaining LLM apps in Azure

• Function calling with OpenAI models
• Creating agents that trigger and consume external APIs
• Integration with Microsoft Graph and Azure Functions
• Building tool-using agents for real-time tasks
• Autogen Overview: Agent orchestration with planners, executors, and tools
• Autogen Overview: Use cases where Autogen shines
• Autogen Overview: Deployment patterns with Azure services

• Observability: monitoring LLM performance and responses
• Managing limitations: latency, cost, and hallucinations
• Security, permissioning, and responsible AI
• Governance models for agentic systems
• Scaling to multi-agent use cases: roadmap and architecture
• Extending Copilot experiences using custom agents

• Build a basic rule-based agent and extend it using Azure OpenAI for dynamic logic
• Create a working RAG application using Azure OpenAI and AI Search to query internal content
• Build an API-integrated agent
• Extend functionality using a starter Microsoft Autogen configuration
• Final project: participants design and present a full-stack agent solution for a selected use case
• Extending Copilot experiences using custom agents