⚙️ Pro Code Agents Path¶
L200 L300 L400
For developers who want full control over their agent architecture. Go beyond managed services and build production-ready agents with open-source frameworks and the Azure AI SDK.
What You'll Build¶
- ✅ Multi-agent systems with AutoGen (orchestrator + specialist agents)
- ✅ Production patterns: retry, circuit breaker, cost control
- ✅ Custom LangChain chains with tool calling
- ✅ End-to-end evaluation pipeline
Path Labs (16 labs, ~1040 min total)¶
| Lab | Title | Level | Cost |
|---|---|---|---|
| Lab 018 | Function Calling & Tool Use | L100 | ✅ GitHub Free |
| Lab 019 | Streaming Responses in Agents | L100 | ✅ GitHub Free |
| Lab 027 | Agent Memory Patterns | L200 | ✅ Free |
| Lab 029 | LangChain & LangGraph Basics | L200 | ✅ Free |
| Lab 035 | Agent Evaluation with Azure AI Eval SDK | L300 | ✅ GitHub Free |
| Lab 036 | Prompt Injection Defense & Agent Security | L300 | ✅ GitHub Free |
| Lab 037 | CI/CD for AI Agents with GitHub Actions | L300 | ✅ GitHub Free |
| Lab 038 | AI Cost Optimization | L300 | ⚠️ Azure |
| Lab 043 | Multimodal Agents with GPT-4o Vision | L300 | ✅ Free |
| Lab 046 | Microsoft Agent 365 — Enterprise Agent Governance | L300 | ⚠️ Azure |
| Lab 057 | Computer-Using Agents — Desktop Automation | L300 | ✅ Free |
| Lab 058 | Browser Automation Agents with OpenAI CUA | L300 | ✅ Free |
| Lab 040 | Production Multi-Agent with AutoGen | L400 | ✅ GitHub Free |
| Lab 044 | Phi-4 + Ollama in Production | L400 | ✅ Free |
| Lab 055 | A2A + MCP Full Stack — Agent Interoperability Capstone | L400 | ✅ Free |
| Lab 064 | Securing MCP at Scale with Azure API Management | L400 | ✅ Free |
Frameworks Covered¶
| Framework | Language | Best for |
|---|---|---|
| AutoGen | Python | Multi-agent conversation orchestration |
| Semantic Kernel | Python / C# | Plugin-based agents, Microsoft stack |
| LangChain | Python / JS | General purpose agent chains |
| Azure AI SDK | Python / C# | Azure-native agent development |