AI-Powered Modular Learning Platform

A modern full-stack web application designed to deliver highly personalized learning experiences. This platform leverages generative AI (Google Gemini 1.5) to dynamically evaluate a learner's existing skills through diagnostic assessments, and automatically generate tailored, modular learning paths based on their performance.

🌟 Key Features

• Diagnostic-First Onboarding: AI-driven assessments that dynamically evaluate user skills before course commencement.
• Personalized Learning Paths: Instead of static linear courses, the platform constructs dynamic pathways linking specific sub-modules tailored to bridge a learner's exact skill gaps.
• Modular Course Architecture: Content is broken down into granular models (Courses -> Modules -> Sub-modules) allowing highly specific content delivery.
• Progress & Performance Tracking: Deep analytics on user engagement, module completion, and learner score evolution.
• Role-Based Access Control & Multi-Tenancy: Robust authentication and organization-based scoping for different users (Learners, Instructors, Admins).

🏗️ Architecture

The platform follows a decoupled client-server architecture, strongly typed with TypeScript across the stack.

Frontend

• Framework: Next.js 16 (App Router) & React 19
• Styling: Tailwind CSS v4
• Animation: Framer Motion for premium micro-interactions and staggered entry animations.
• Icons: Lucide React
• Structure:
  • /app/onboarding & /app/login: Authentication and initial diagnostic workflows.
  • /app/dashboard: Main user interface displaying the personalized learning trajectory.
  • /src/services: Decoupled API client services (Auth, Course, Learner, Assessment).

Backend

• Framework: Node.js with Express v5
• Language: TypeScript
• Database: MongoDB with Mongoose ODM
• AI Integration: Google Generative AI (@google/generative-ai) via a dedicated aiGenerator utility to synthesize JSON-structured assessments and curriculum paths.
• Authentication: JSON Web Tokens (JWT) & bcryptjs for secure password hashing.
• Mailing: Nodemailer for transactional email logs.

Database Schema (Core Models)

• User & Organization: Core identity and multi-tenant scoping.
• Course & Module: The hierarchical structure of the learning content.
• Assessment & LearnerScore: Diagnostic evaluations and the resulting normalized skill metrics.
• PersonalizedPath: The AI-generated mapping of Modules assigned to a specific User.
• Enrollment & ProgressTracking: State management for a user's journey through their assigned path.

🚀 Getting Started

Prerequisites

• Node.js (v20+)
• MongoDB Instance (Local or Atlas)
• Google Gemini API Key

Installation

1. Clone the repository

2. Setup Backend:

   cd backend
   npm install

   Create a .env file in the backend directory:

   PORT=5000
   MONGO_URI=your_mongodb_connection_string
   JWT_SECRET=your_jwt_secret
   GEMINI_API_KEY=your_gemini_api_key

   Run the backend dev server:

   npm run dev

3. Setup Frontend:

   cd frontend
   npm install

   Create a .env.local file in the frontend directory:

   NEXT_PUBLIC_API_URL=http://localhost:5000/api

   Run the frontend dev server:

   npm run dev

🧠 AI Assessment Pipeline Workflow

1. Trigger: Learner initiates the onboarding sequence.
2. Generation: Backend assessmentController calls the aiGenerator service. Gemini 1.5 analyzes the target course domains and returns a structured JSON payload containing diagnostic questions.
3. Evaluation: Learner submits answers; the system evaluates and updates the LearnerScore document.
4. Path Synthesis: Based on the scores, the AI determines which modules can be skipped and which are mandatory, persisting this into a PersonalizedPath.