Architecture¶

High-Level Architecture

This document provides an overview of the IncLearn system architecture.

High-Level Architecture¶

IncLearn follows a modern web application architecture:

┌─────────────┐     ┌─────────────┐     ┌─────────────┐
│   Web UI    │────▶│   Server    │────▶│  Database   │
│  (Next.js)  │     │ (FastAPI)   │     │ (CosmosDB)  │
└─────────────┘     └─────────────┘     └─────────────┘
                           │
                           ▼
                    ┌─────────────┐
                    │ Azure AI    │
                    │ Services    │
                    └─────────────┘

Components¶

Web Frontend¶

Framework: Next.js (React)
Language: TypeScript
Styling: Material UI
State Management: Jotai

Backend Server¶

Framework: FastAPI
Language: Python 3.13+
ORM: SQLAlchemy
Authentication: OAuth2 / JWT

Database¶

Type: Azure CosmosDB for PostgreSQL
Purpose: Store users, courses, organizations, teams

AI Services¶

Azure AI Foundry - Content evaluation
Azure Cognitive Services - Captioning, transcription
Azure AI Agent - Content processing and accessibility evaluation

Key Features¶

Accessibility Automation¶

User publishes content
Server triggers content accessibility service, which invokes Azure AI Agent
AI evaluates and generates alt text, captions, transcripts
Results stored in database
Accessibility report stored in database with metadata

Content Management¶

Course structure: Organization → Course → Module → Chapter
Rich content types with metadata

User Management¶

Role-based access control (RBAC)
Organization and team hierarchy
Learner profiles with preferences

API¶

The backend exposes a RESTful API with OpenAPI documentation.

See API for details.

Security¶

JWT-based authentication
Role-based access control
Input validation
SQL injection prevention
XSS protection

Scalability¶

Stateless backend
Database connection pooling
CDN for static assets
AI services scale automatically