Reviews 4,2
Reviews 4,5
100% fundable

AI-3026 Develop AI agents on Azure

AI agents act autonomously – they plan steps, use tools, and respond adaptively to their environment. This training teaches you how to design, build, and operate such agents on the Azure platform.

Key topics

  • Fundamentals of AI agent architecture and autonomous reasoning patterns
  • Building agents with Azure AI Foundry and Semantic Kernel
  • Connecting tools, APIs, and external data sources to your agents
  • Designing and orchestrating multi-agent systems for complex tasks
  • Monitoring, tracing, and optimizing agents in production environments

Prerequisite

Familiarity with Python and a basic understanding of cloud services is expected. Prior experience with Azure is helpful but not required.

Target group

This training is designed for developers and AI engineers who want to implement intelligent, autonomous systems on Azure and integrate them into real-world workflows.

Course content
  • Understand the concept and architecture of autonomous AI agents
  • Explore reasoning patterns: ReAct, chain-of-thought, and tool use
  • Distinguish agent types and their use cases in practice
  • Navigate Azure AI Foundry and configure your project environment
  • Select and deploy models suited for agent-based applications
  • Set up prompt flows and tools within the Foundry workspace
  • Implement agents using Semantic Kernel plugins and planners
  • Define skills and connect them to agent reasoning pipelines
  • Integrate memory and context management into your agent logic
  • Connect REST APIs, databases, and search services to your agent
  • Design tool call flows and handle responses within the agent loop
  • Secure and manage external connections in production-ready setups
  • Design systems where multiple specialized agents collaborate on tasks
  • Implement coordination patterns and handoff logic between agents
  • Handle conflicts, retries, and failure scenarios in agent networks
  • Implement short-term and long-term memory for agent continuity
  • Manage conversation context and session state across interactions
  • Use vector stores and retrieval-augmented generation within agents
  • Apply security principles to agent inputs, outputs, and tool access
  • Implement guardrails and content filters for safe agent behavior
  • Align agent systems with responsible AI guidelines and compliance requirements
  • Set up logging, metrics, and tracing for agents running on Azure
  • Catch failure scenarios and implement fallback mechanisms
  • Optimize agent performance iteratively during live operation
  • Implement a complete AI agent from concept through deployment
  • Work through concrete real-world scenarios based on practical requirements
  • Test, iterate, and refine the agent using feedback and observability data

Your Personal Path to the Right Continuing Education Program—We Provide Personalized Guidance

Our experienced educational advisors will guide you from your very first question all the way to the start of your continuing education program.