Creating Scalable AI Agents on Azure Using Microsoft Foundry

• Introduction to Agentic AI on Azure
  • What are AI agents?
  • Single-agent vs multi-agent systems
  • Tool usage & function calling
  • Planning and reasoning loops
  • Enterprise use cases for agents
• Azure AI Foundry Overview
  • Azure AI Foundry architecture
  • Model catalog & deployment lifecycle
  • Azure OpenAI integration
  • Prompt Flow fundamentals
  • Azure AI Studio capabilities
• Building First Azure Agent
  • Creating AI Foundry project
  • Deploying foundation model
  • Designing structured prompts
  • Implementing function calling
  • Testing agent workflows
• Integrating Enterprise Tools
  • Connecting REST APIs
  • Secure credential management
  • Managed identity integration
  • Logging and diagnostics
• Hands-on Labs
  • Deploy Azure OpenAI model
  • Build tool-calling agent
  • Create Prompt Flow workflow
  • Integrate external API
  • Analyze logs and outputs

• Multi-Agent Architecture on Azure
  • Supervisor-agent pattern
  • Planner-executor model
  • Role-based agent specialization
  • State management strategies
  • Context persistence
• Retrieval-Augmented Generation (RAG)
  • Azure AI Search integration
  • Embeddings generation
  • Chunking & indexing strategy
  • Hybrid search patterns
  • Hallucination mitigation
• Agent Orchestration & Workflow Design
  • Conditional branching
  • Task delegation
  • Retry & fallback logic
  • Event-driven orchestration
  • Monitoring agent decisions
• Observability & Evaluation
  • Agent performance metrics
  • Latency tracking
  • Token usage optimization
  • Automated evaluation workflows
  • Human-in-the-loop validation
• Hands-on Labs
  • Build multi-agent workflow
  • Integrate RAG pipeline
  • Implement supervisor agent
  • Add memory persistence
  • Measure latency & token usage

• Production Architecture for Scalable Agents
  • API-based deployment
  • Containerization (Azure Container Apps / AKS)
  • Serverless deployment patterns
  • Scaling concurrent sessions
  • Load balancing strategies
• Security & Governance Framework
  • Role-based access control (RBAC)
  • Private endpoints
  • Key Vault integration
  • Data privacy & PII protection
  • Responsible AI guardrails
• Performance & Cost Optimization
  • Token optimization strategies
  • Caching layers
  • Async processing
  • Throughput tuning
  • Budget monitoring
• Enterprise AI Operating Model
  • AI Center of Excellence
  • Governance committees
  • Risk classification framework
  • Change management strategy
  • ROI measurement
• Capstone Project: Scalable Enterprise Agent System
  • Define real-world enterprise use case
  • Design multi-agent architecture
  • Integrate RAG + enterprise APIs
  • Deploy using Azure scalable infrastructure
  • Present architecture trade-offs & governance plan

Upon completion, participants will be able to:

• Design scalable AI agents using Azure AI Foundry
• Build multi-agent orchestrated systems
• Integrate RAG pipelines and enterprise APIs securely
• Deploy and monitor production-grade agent systems
• Implement governance, security, and Responsible AI controls
• Optimize cost, latency, and performance