Demo: The Complete Application

Duration: 20 minutes

What You’ll See

In this demonstration, you’ll explore a complete AI-powered personalised worked example generator. The application demonstrates how AI can create customized learning materials tailored to individual student contexts.

What makes this interesting:

  • Generates examples across 3 different domains (Programming, Health Sciences, Agronomy)
  • Personalizes content based on student interests, goals, and background
  • Creates unique worked examples on demand

Try the Interactive Demo

Explore the complete application below. You can interact with it directly or open it in a new window for a larger view.

Open demo in new window

How to Explore the Demo

Step 1: Examine the Interface

Notice the clean, simple layout:

  • Learner profile section (name, domain, interests, goals)
  • Concept selection dropdown
  • Generate button
  • Display area for the personalized example

This interface was built using Marimo, a reactive Python notebook framework.

Step 2: Try Different Learner Profiles

Example 1: Programming Domain

Create a profile like this:

Name: Sarah
Domain: Programming (Python)
Specific interest: Web development
Hobby/passion: Baking
Goal: Build a recipe sharing website
Level: Beginner

Select a concept (e.g., “For Loops”) and click Generate.

What to notice:

  • The example uses baking contexts (recipes, ingredients)
  • Code examples relate to her recipe website goal
  • The worked solution is complete and step-by-step
  • Practice suggestions continue the baking theme

Example 2: Health Sciences Domain

Try a different profile:

Name: Marcus
Domain: Health Sciences (Statistics)
Specific interest: Sports performance
Hobby/passion: Running
Goal: Improve marathon time
Level: Intermediate

Select “Linear Regression” and generate.

What to notice:

  • Same tool, completely different domain
  • Examples now use running and race data
  • Statistical concepts applied to sports performance
  • Personalization adapts to the new context

Example 3: Agronomy Domain

If you have time, try:

Name: Elena
Domain: Agronomy
Specific interest: Coffee farming
Hobby/passion: Sustainability
Goal: Increase yield sustainably
Level: Beginner

This demonstrates the breadth: 3 domains, 16 concepts total.

What Makes This Powerful?

Why This Matters

The traditional challenge:

Creating personalized learning materials is time-intensive. Writing a single customized worked example takes 15-30 minutes. For a class of 20 students learning 16 different concepts, you’d need to create 320 unique examples. This is simply not feasible for most educators.

The AI-powered solution:

The application generates personalized examples in 30-60 seconds. Each student gets content tailored to their interests and goals, and the approach scales to any number of learners.

How It Works

The application follows a simple three-step process:

  1. Collect context: Learner fills in their profile (interests, goals, background)
  2. Generate example: AI creates a worked example tailored to that specific learner
  3. Display result: The personalized example appears with complete solution steps

Behind the scenes, the AI uses the learner’s context to weave familiar scenarios into the teaching content, making abstract concepts more concrete and relatable.

Domains and Concepts

Current coverage:

  • Programming (Python): For loops, list comprehensions, dictionaries, functions, string formatting (and more)
  • Health Sciences (Statistics): Correlation, mean/SD, t-tests, confidence intervals, linear regression (and more)
  • Agronomy: Yield prediction, NPK optimization, growing degree days, water efficiency

The pattern can extend to any domain where worked examples support learning.

Reflection Questions

Think About

Consider these questions as you explore the demo:

  1. What aspects of the personalization impressed you most?
  2. How could personalized worked examples benefit YOUR students?
  3. What domain or subject area would be relevant for your teaching context?
  4. What types of student interests or goals could you leverage for personalization?

Key insight: AI-powered tools can scale individualized instruction in ways that were previously impractical. The pattern demonstrated here (structured data + AI generation + interactive UI) can apply across many educational contexts.

Pedagogical Foundation

This tool is grounded in Cognitive Load Theory research:

Worked Example Effect: Novice learners benefit more from studying complete solutions than from problem-solving alone (effect size 0.52).

Personalization Principle: Familiar contexts reduce extraneous cognitive load, allowing learners to focus on the target concept rather than decoding unfamiliar scenarios.

By combining these principles with AI generation, we can provide:

  • High-quality worked examples (supporting novice learning)
  • Personalized to individual contexts (reducing cognitive load)
  • At scale (reaching all learners, not just a privileged few)
Back to top