NEXUS-LEARN

NEXUS-LEARN is a cloud-native, adaptive learning platform built to improve student outcomes while reducing institutional overhead. It combines real-time learner profiling, AI-assisted personalization, and an administrative command center in a single React + Firebase application.

LIVE WEBSITE URL : https://project-nexus-learn.vercel.app/

Executive Overview

The Problem

Education technology is split between:

• Expensive enterprise LMS platforms with heavy infrastructure and licensing overhead (often quoted on a per-student basis).
• Generic one-size-fits-all systems that do not adapt to individual learning behavior.

The result is predictable: educators can spend as much as 40% of their time on administrative work, while students are left with static learning experiences that ignore pace, context, and cognitive state.

The NEXUS-LEARN Solution

NEXUS-LEARN addresses three core challenges:

1. Cost efficiency with serverless architecture: Firebase Auth + Firestore remove traditional backend infrastructure burden and support usage-based scaling, with a target of up to 80% lower platform cost compared with traditional enterprise LMS setups.
2. Personalization at scale: A React dashboard continuously adapts learner-facing insights and recommendations based on profile and behavior signals.
3. Administrative intelligence: An institutional panel aggregates learner risk and fatigue signals to support faster intervention and better academic oversight.

Technical Innovation

Instead of relying on heavyweight training pipelines, NEXUS-LEARN uses an intelligent, practical architecture:

• Deterministic local modeling (computeTwin) for immediate baseline insights.
• Optional Gemini augmentation for deeper narrative reasoning and personalization.
• Firestore-backed document data model for fast, index-friendly profile and cohort queries.

This keeps the platform fast, extensible, and deployable now.

Core Product Modules

SENTINEL

Cognitive load and fatigue intelligence engine.

• Fatigue scoring (0-100) from wellbeing and workload signals.
• 72-hour cognitive forecast bands (green/amber/red).
• Intervention cueing (schedule nudge, rest recommendation, advisor alert, pastoral flag).

IRIS

Adaptive reading and comprehension engine.

• Dynamic typography and reading-surface adaptation.
• Text complexity analysis and reading profile adjustments.
• Optional Gemini Q&A for student-entered notes/questions.

APOLLO

Academic risk and institutional visibility engine.

• Dropout risk tiering (low/moderate/high/critical).
• Factor-level risk contributors.
• Cohort-level summaries for advisor/admin review.

📸 Screenshots

What Is Implemented Today

• Branded landing and onboarding experience.
• 4-step student profile capture (identity, wellbeing, learning, academics).
• Firebase Authentication (email/password).
• Firestore-backed user persistence.
• Student dashboard with tabs: Overview, Sentinel, Iris, Apollo.
• AI-augmented digital twin and recommendations (when Gemini key is configured).
• Deterministic fallback mode when Gemini is unavailable.
• Institutional cohort console with peer visibility and intervention indicators.
• Admin panel with:
  • platform metrics,
  • student search/listing,
  • profile edit/delete flows,
  • recent activity view.

Tech Stack

• Frontend: React 19, React DOM, Create React App (react-scripts)
• Data/Auth: Firebase Auth + Cloud Firestore
• AI: Gemini 2.5 Flash API (optional but supported)
• Testing: React Testing Library + Jest DOM
• Styling: In-component design system + injected global CSS tokens

Architecture (High Level)

Student Input/Profile
        |
        v
Deterministic Twin Engine (SENTINEL + IRIS + APOLLO baseline)
        |
        +--> Optional Gemini Enrichment (analysis, recs, Q&A)
        |
        v
Adaptive UI (student + institutional + admin views)
        |
        v
Firebase (Auth + Firestore user documents + session state)

Repository Structure

nexus-learn/
|-- docs/
|   `-- images/
|-- public/
|   `-- index.html
|-- src/
|   |-- NexusLearnDashboard.jsx
|   |-- firebase.js
|   `-- index.js
|-- .env
|-- .gitignore
|-- package.json
`-- package-lock.json

Data Model (Firestore)

User documents are stored in the users collection (document key: lowercased email).

{
  "uid": "u_...",
  "email": "student@example.com",
  "fullName": "Student Name",
  "cohort": "B.Tech Computer Science & Engineering",
  "studentId": "AUTOID",
  "institution": "University Name",
  "yearOfStudy": "2",
  "profile": {
    "sleepHours": "7",
    "stressLevel": "5",
    "studyHoursPerDay": "4",
    "exercisePerWeek": "3",
    "screenTimeHours": "6",
    "caffeine": "2",
    "lastRestDays": "1",
    "hydration": "moderate",
    "breakFrequency": "2",
    "mentalHealthRating": "6",
    "socialEngagement": "3",
    "deadlineAnxiety": "moderate",
    "workingStatus": "student",
    "readingSpeed": "250",
    "motivationLevel": "7",
    "courseDifficulty": "3",
    "missedDeadlines": "0",
    "attendanceRate": "85",
    "gpaLast": "3.0"
  },
  "createdAt": "ISO_DATE",
  "lastUpdated": "ISO_DATE",
  "loginHistory": ["ISO_DATE"]
}

Local Setup

Prerequisites

• Node.js 18+ (recommended)
• npm 9+ (recommended)
• Firebase project (Auth + Firestore enabled)
• Gemini API key (optional but recommended)

1. Install dependencies

npm install

2. Configure environment variables

Create a .env file in the project root:

REACT_APP_GEMINI_API_KEY=your_gemini_key

REACT_APP_FIREBASE_API_KEY=your_firebase_api_key
REACT_APP_FIREBASE_AUTH_DOMAIN=your_project.firebaseapp.com
REACT_APP_FIREBASE_PROJECT_ID=your_project_id
REACT_APP_FIREBASE_STORAGE_BUCKET=your_project.appspot.com
REACT_APP_FIREBASE_MESSAGING_SENDER_ID=your_sender_id
REACT_APP_FIREBASE_APP_ID=your_app_id
REACT_APP_FIREBASE_MEASUREMENT_ID=your_measurement_id

3. Run in development

npm start

Runs at http://localhost:3000.

4. Build for production

npm run build

Available Scripts

• npm start: Run local dev server
• npm run build: Create production build
• npm test: Run tests
• npm run eject: Eject CRA config

Security Notes (Important)

• Move admin credentials out of source code and into secure auth/role management before production.
• Rotate any exposed API keys immediately if they were committed.
• .env is gitignored — never commit it. Use .env.example as a reference template.
• Enforce Firebase Security Rules for least-privilege access.

Deployment

NEXUS-LEARN can be deployed as a static React bundle with managed backend services:

• Frontend hosting: Firebase Hosting, Vercel, Netlify, or any static host.
• Backend services: Firebase Auth + Firestore.
• AI layer: Gemini API via REACT_APP_GEMINI_API_KEY.

Roadmap

• Split monolithic dashboard into domain modules/components.
• Replace hardcoded admin shortcut with role-based Firebase authorization.
• Add audit logging and stronger operational analytics.
• Extend adaptive content modules (quiz workflows, media-based learning flows).

License

This project is licensed under the MIT License. Copyright (c) 2026 Rohit Jain.

Author

Rohit Jain (B.Tech CSE) LinkedIn: https://www.linkedin.com/in/546-rohit-jain

Debdatta Panda LinkedIn: https://www.linkedin.com/in/myself-debdatta-194a822b1

Soumyasri Mohapatra LinkedIn: https://www.linkedin.com/in/soumyasri-mohapatra-99211625b

Ritisha Sahoo LinkedIn: https://www.linkedin.com/in/ritisha-sahoo-67a046364

Sahil Kumar sahoo LinkedIn: https://www.linkedin.com/in/sahil-kumar-sahoo-4055b7368