Create an agent that remembers user preferences (favorite color, hobby, pet name) and greets the user with their details on each new conversation.
Use FileMemory for persistence
Store at least 3 user attributes
Retrieve and display them on subsequent visits
Lab 2: Tool-Powered Agent
Build an agent that can perform file operations: list files, read file contents, and write to files. The agent should decide which operation to perform based on user intent.
Three tools: list_files, read_file, write_file
Proper error handling for missing files
Test with multi-step requests like "Read notes.txt, then create a summary.txt"
Lab 3: Multi-Agent Q&A System
Build a multi-agent system where a triage agent routes questions to a technical documentation agent or a policy agent depending on the topic.
Triage agent with routing rules
Doc agent loads from a knowledge base
Policy agent applies rule-based responses
Context passed between agents
Lab 4: Production Deployment
Deploy one of your agents to Azure Container Apps with full observability:
Containerize your agent with Docker
Push to Azure Container Registry
Deploy to Azure Container Apps
Configure Application Insights
Test the deployed agent via REST API
Study Plan
Week
Topic
Activities
1
Foundations & Setup
Read Ch 1-2, complete setup, Lab 1
2
First Agent & Capabilities
Read Ch 3-4, build first agent, add memory
3
Tools & Extensions
Read Ch 5, build tool suite, Lab 2
4
Orchestration
Read Ch 6, multi-agent patterns, Lab 3
5
Governance & Deployment
Read Ch 7, deploy to Azure, Lab 4
✏️ Exercise: Choose one of the labs above and complete it end-to-end. Write a short reflection (3-5 sentences) on what you learned, what was challenging, and how you would apply it to a real-world project.