Notebook Development Skill (Teaching Labs)

Overview

This skill helps create well-structured Jupyter notebooks for teaching Northwestern MS AI students about Azure services and agentic AI.

Target Audience

•Who: Graduate students in MS AI program
•Background: Smart industry professionals, strong coding skills
•Tone: Professional, educational, explains why not just how
•Level: Intermediate to advanced Python, new to Azure/AI Foundry

Notebook Structure Template

code

1. Title & Overview (Markdown)
   - Lab title
   - Learning objectives
   - Prerequisites
   - Architecture diagram

2. Setup & Prerequisites (Code + Markdown)
   - Import checks
   - Tool verification
   - Environment setup

3. Core Concepts (Markdown)
   - Explain the theory
   - Architecture diagrams
   - Key terminology tables

4. Step-by-Step Implementation (Code + Markdown)
   - One concept per section
   - Code cells with comments
   - Explanatory markdown between code

5. Testing & Verification (Code)
   - Test the implementation
   - Show expected outputs

6. Summary (Markdown)
   - Key takeaways table
   - Next steps
   - Additional resources

Writing Style Guidelines

Markdown Cells

markdown

## Step X: Clear Action Title

**Why this matters:** Brief explanation of the concept's importance.

### Key Concepts

| Concept | Description |
|---------|-------------|
| Term 1  | Clear definition |
| Term 2  | Clear definition |

### Architecture

\`\`\`
┌─────────────┐     ┌─────────────┐
│  Component  │────▶│  Component  │
└─────────────┘     └─────────────┘
\`\`\`

Now let's implement this:

Code Cells

python

# Clear section comment explaining what this code does
# Include inline comments for non-obvious logic

def example_function(data: list[float]) -> dict:
    """
    Brief description of what the function does.
    
    Args:
        data: List of numbers to process
        
    Returns:
        Dictionary with computed statistics
    """
    # Validate input
    if not data:
        raise ValueError("Data cannot be empty")
    
    # Compute statistics
    result = {
        "count": len(data),
        "mean": sum(data) / len(data),
    }
    
    return result

# Test the function
sample_data = [10, 20, 30, 40, 50]
result = example_function(sample_data)
print(f"✅ Analysis complete: {result}")

ASCII Art Diagrams

Use consistent styles for architecture diagrams:

Flow Diagram

code

Input ───▶ Process ───▶ Output

Component Diagram

code

┌─────────────────┐     ┌─────────────────┐
│   Component A   │────▶│   Component B   │
│   (description) │◀────│   (description) │
└─────────────────┘     └─────────────────┘

Layered Architecture

code

┌─────────────────────────────────────────┐
│            Presentation Layer            │
├─────────────────────────────────────────┤
│             Business Logic              │
├─────────────────────────────────────────┤
│              Data Layer                 │
└─────────────────────────────────────────┘

Emoji Conventions

Use emojis consistently:

Emoji	Meaning
✅	Success/Complete
❌	Error/Failure
⚠️	Warning/Caution
📝	Note/Documentation
🔧	Configuration/Setup
🎉	Celebration/Milestone
📡	Network/API
🔒	Security
💡	Tip/Insight

Tables for Key Concepts

markdown

| Feature | Description | Example |
|---------|-------------|---------|
| Feature 1 | Clear description | `code example` |
| Feature 2 | Clear description | `code example` |

Code Output Formatting

python

# Print structured output
print("=" * 50)
print("📊 Results Summary")
print("=" * 50)
print(f"  • Count: {result['count']}")
print(f"  • Mean:  {result['mean']:.2f}")
print("=" * 50)

Prerequisites Check Pattern

python

import subprocess
import shutil

def check_tool(name: str, command: list) -> bool:
    """Check if a tool is installed."""
    try:
        result = subprocess.run(command, capture_output=True, text=True, timeout=10)
        version = result.stdout.strip().split('\n')[0]
        print(f"✅ {name}: {version}")
        return True
    except (subprocess.SubprocessError, FileNotFoundError):
        print(f"❌ {name}: Not installed")
        return False

print("Checking required tools...\n")
tools = [
    ("Python", ["python", "--version"]),
    ("Azure CLI", ["az", "--version"]),
]
all_ok = all(check_tool(name, cmd) for name, cmd in tools)

if all_ok:
    print("\n🎉 All tools installed!")
else:
    print("\n⚠️ Some tools missing. See installation instructions above.")

Summary Section Template

markdown

## 🎯 Summary

### What You Built
Brief description of the complete implementation.

### Key Takeaways

| Concept | Implementation |
|---------|----------------|
| Concept 1 | How it was implemented |
| Concept 2 | How it was implemented |

### Next Steps
- Link to next lab
- Advanced topics to explore
- Additional resources

## 📚 Resources
- [Official Documentation](url)
- [Tutorial](url)