The Architecture of Learning

Introduction

Every great platform is built on a simple architecture.

Not a software architecture.

A conceptual architecture.

It defines the language of the platform before a single line of code is written.

It tells us what exists.

Why it exists.

How each part relates to every other part.

This chapter defines the architecture of Maigie.

Every future feature, screen, API, database model and AI capability should fit naturally within this architecture.

If something cannot find its place here, it probably does not belong in the product.

The Philosophy of the Architecture

Learning is not a collection of disconnected features.

Learning is an ecosystem.

Within every ecosystem, each part has one clear responsibility.

Learners pursue growth.

Learning Spaces organize communities.

Classrooms organize structured learning.

Courses organize knowledge.

Learning Intelligence continuously improves the experience.

When every object has one responsibility, the platform remains understandable as it grows.

The Core Domain Model

The platform is composed of five primary domain objects.

Learner
├── Personal Learning
├── Learning Spaces
│      │
│      ├── Classrooms
│      │      │
│      │      ├── Discussions
│      │      ├── Sessions
│      │      ├── Assignments
│      │      ├── Announcements
│      │      ├── Progress
│      │      └── Assigned Courses
│      │
│      └── Members
│
├── Courses
│      ├── Modules
│      ├── Resources
│      ├── Assessments
│      └── Learning Objectives
│
└── Learning Intelligence

These objects form the foundation of the platform.

Every other capability extends one of them.

Learner

The learner is the centre of Maigie.

Everything exists to help the learner move from potential to achievement.

Every learner owns a Personal Learning environment.

Every learner may belong to multiple Learning Spaces.

Every learner builds relationships with knowledge, educators and peers.

The learner is never passive.

The learner is an active participant in their own transformation.

Personal Learning

Personal Learning belongs exclusively to one learner.

It is independent of every Learning Space.

Within Personal Learning, learners may:

• Chat with Maigie
• Prepare for examinations
• Learn professional skills
• Generate academic documents
• Create personal notes
• Build flashcards
• Plan study schedules
• Track goals
• Measure progress
• Save resources

Everything here exists to strengthen the individual learner.

Learning Spaces

Learning Spaces organize collaborative learning.

A Learning Space represents a long-lived educational environment.

Examples include:

• A university department
• A tutor’s cohort
• A bootcamp
• A study group
• A company training programme

Learning Spaces own:

• Members
• Roles
• Permissions
• Analytics
• Settings
• Classrooms

A Learning Space does not own knowledge.

It organizes the people who use knowledge.

Classrooms

Classrooms are the operational units of collaborative learning.

Every Classroom exists around a shared learning objective.

Examples include:

• CSC401 – Software Engineering
• AWS Solutions Architect
• Final Year Project
• IELTS Preparation

Each Classroom may contain:

• Discussions
• Learning Sessions
• Voice Sessions
• Assignments
• Announcements
• Classroom Resources
• Progress
• AI assistance
• Assigned Courses

The Classroom is where learners, educators and knowledge come together around a shared objective.

Courses

Courses represent structured knowledge.

A Course is reusable.

It is not owned by any Classroom.

Instead, Classrooms assign Courses.

A Course contains:

• Learning objectives
• Modules
• Concepts
• Resources
• Assessments
• Recommended learning sequence

As learners interact with Courses, Maigie identifies opportunities to improve them.

Educators may contribute additional resources, examples or assessments based on recommendations from Learning Intelligence.

Courses therefore become continuously stronger over time.

Learning Intelligence

Learning Intelligence powers the entire platform.

It is not another feature.

It is infrastructure.

Learning Intelligence observes every domain object.

It remembers.

Reasons.

Connects.

Recommends.

Schedules.

Supports.

Improves.

It owns no data.

Instead, it creates value from the relationships between existing data.

Users should never think:

“I am using the AI.”

Instead they should naturally say:

“Maigie helped me.”

Architectural Laws

Every future capability should respect the following principles.

Everything begins with the learner.

Every object ultimately exists to help a learner grow.

Every learner owns Personal Learning.

This remains true regardless of whether they participate in collaborative learning.

Learning Spaces organize people.

They do not own knowledge.

Courses are reusable.

Knowledge should be shared whenever appropriate rather than duplicated.

Classrooms organize learning.

Learning Sessions create learning.

Persistent environments and learning moments are distinct concepts.

Learning Intelligence owns nothing.

It observes everything.

The Responsibility Test

Whenever a new feature is proposed, ask three questions.

1. Which domain object owns this capability?
2. Does it strengthen that object’s responsibility?
3. Does it move the learner closer to transformation?

If the answers are unclear, the architecture is telling us that we have not yet fully understood the feature.

The correct response is not to write more code.

It is to improve the model.

Closing Thoughts

Software becomes difficult when its architecture reflects features.

Software becomes enduring when its architecture reflects reality.

Maigie is not organised around screens.

It is organised around learning itself.

As long as this architecture remains clear, the platform can evolve for many years without losing its identity.