Skip to content

next-steps.md

Latest commit

 

History

History
179 lines (124 loc) · 4.92 KB

next-steps.md

File metadata and controls

179 lines (124 loc) · 4.92 KB

4️⃣ Next Steps

Congratulations! 🎉 You've built and deployed your first AI agent on Azure.

🎯 What You've Accomplished

  • ✅ Set up your development environment
  • ✅ Built an AI agent with the Microsoft Agent Framework
  • ✅ Deployed to Azure with proper security
  • ✅ Learned the fundamentals of agentic AI on Azure

🗺️ Where to Go From Here

📚 Dive Deeper into This Repo

Topic Guide Description
Architecture Architecture Best Practices Reference architectures, AI Landing Zones
Agent Patterns Agent Development Multi-agent workflows, tools, error handling
Security Security Best Practices Private endpoints, RBAC, governance
Deployment Production Deployment CI/CD, IaC, monitoring
SDKs SDK Reference Agent Framework, Foundry SDK, Copilot SDK

🔧 Advanced Topics

Multi-Agent Workflows

Build systems where multiple agents collaborate:

from agent_framework.workflows import Workflow, Graph

# Create specialized agents
planner = create_agent("Planner", "Break down complex tasks")
researcher = create_agent("Researcher", "Find information")
writer = create_agent("Writer", "Create content")

# Connect them in a workflow
workflow = Workflow.from_graph(
    Graph()
    .add_node("plan", planner)
    .add_node("research", researcher)
    .add_node("write", writer)
    .add_edge("plan", "research")
    .add_edge("research", "write")
)

result = await workflow.run("Create a report on Azure AI")

📖 See full example

Add Observability

Monitor your agents in production:

from microsoft_agents_a365.observability.core.config import configure
from microsoft_agents_a365.observability.extensions.agentframework.trace_instrumentor import (
    AgentFrameworkInstrumentor,
)

# Configure observability
configure(service_name="my-agent", service_namespace="production")

# Enable auto-instrumentation
AgentFrameworkInstrumentor().instrument()

📖 Agent 365 Observability Docs

Deploy an AI Landing Zone

For enterprise-scale deployments:

git clone https://github.com/Azure/AI-Landing-Zones.git
cd AI-Landing-Zones/bicep
azd up

📖 AI Landing Zones Repository

Building Developer Tools? Try Copilot SDK

If you're building developer-focused tools (CLIs, IDE extensions, code review bots), the GitHub Copilot SDK might be a better fit:

import { CopilotClient, defineTool } from "@github/copilot-sdk";

// Define a custom tool
const reviewCode = defineTool("review_code", {
    description: "Review code for issues",
    parameters: {
        type: "object",
        properties: {
            code: { type: "string", description: "The code to review" },
        },
        required: ["code"],
    },
    handler: async ({ code }) => {
        // Your review logic here
        return { issues: [], suggestions: ["Add type hints"] };
    },
});

const client = new CopilotClient();
const session = await client.createSession({
    model: "gpt-4.1",
    tools: [reviewCode],
});

const response = await session.sendAndWait({ 
    prompt: "Review this function for issues..." 
});
console.log(response?.data.content);

await client.stop();

📖 Full Copilot SDK Tutorial | When to Use What

📖 Official Microsoft Resources

Resource Description
Agent Framework Docs Official documentation
Azure AI Foundry Platform documentation
Azure Architecture Center Reference architectures
Cloud Adoption Framework Enterprise guidance

💬 Join the Community

Platform Link
🎮 Discord Join Code to Cloud
🐙 GitHub Code to Cloud Org
💬 Discussions Ask Questions

🤝 Contribute

Found something to improve? We'd love your help!

  1. Fork the repo
  2. Create a feature branch
  3. Make your changes
  4. Submit a pull request

📖 Contributing Guide

🚀 Happy Building!

You're now ready to build amazing AI agents on Azure.

← Back to Main README