Project Nexus

Intelligent Academic Workflow Automation System

Project Nexus transforms a messy PDF syllabus into a living, breathing study plan. It doesn't just tell you what to study—it executes your schedule, finds resources, and tracks your mastery in real-time.

🎯 Overview

Project Nexus is a comprehensive academic workflow system that takes syllabus input and creates an intelligent, automated study experience across four phases:

1. Context Phase - Extract structured data from messy PDFs
2. Agentic Phase - Automatically schedule and execute study plans
3. Active Phase - Practice tests and active recall during study
4. Mastery Phase - 3D visualization of knowledge mastery

✨ Key Features

Context Phase (Ingestion)

• PDF Parsing: Extracts deadlines, weightages, topics, and required readings from messy PDFs
• Natural Language Processing: Parse text input like "I want to pass CS101 at UCSB; here is my midterm date"
• Privacy-First: Uses local LLMs (Ollama) so student data never leaves the machine

Agentic Phase (Orchestration)

• Google Calendar Sync: Finds "dead air" (gaps between classes/gym/sleep) automatically
• Task Insertion: Creates specific study blocks with topics and resource links
• Resource Discovery: Finds relevant YouTube videos, Coursera courses, and EdX modules
• Auto-Enrollment: Uses Selenium to automatically enroll in mentioned courses (HTML-as-truth navigation)

Active Phase (Socratic Loop)

• Practice Test Generator: Builds simulated midterms/finals based on your specific syllabus
• Active Recall Chrome Extension: Interrupts YouTube study sessions with contextual questions
• Adaptive Questions: Questions generated based on what you're actually watching

Mastery Phase (Visual Flex)

• 3D Concept Flow Chart: Interactive 3D graph showing knowledge mastery
• Color-Coded Nodes: Green (mastered), Yellow (learning), Orange (started), Red (untouched)
• Real-Time Progress Tracking: Update mastery as you learn and see the map evolve

🏗️ Architecture

Project-Nexus/
├── backend/                 # Flask API server
│   ├── context_phase/      # Syllabus parsing and extraction
│   ├── agentic_phase/      # Calendar sync, task management, course enrollment
│   ├── active_phase/       # Practice test generation
│   ├── mastery_phase/      # 3D visualization
│   └── app.py              # Main Flask application
├── frontend/               # React web application
│   ├── src/
│   │   ├── components/     # React components
│   │   └── App.js          # Main React app
│   └── package.json
├── chrome-extension/       # Active recall Chrome extension
│   ├── manifest.json
│   ├── content.js          # YouTube video detection and questions
│   └── popup.html
└── README.md

🚀 Quick Start

Prerequisites

• Python 3.8+
• Node.js 16+
• Google Chrome
• ChromeDriver (for Selenium)
• Ollama (optional, for local LLM)

Installation

1. Clone the repository

git clone <repository-url>
cd Project-Nexus

2. Set up the backend

cd backend
python -m venv venv
source venv/bin/activate  # Windows: venv\Scripts\activate
pip install -r requirements.txt
cp .env.example .env  # Edit with your API keys
python app.py

3. Set up the frontend

cd frontend
npm install
npm start

4. Install Chrome Extension

• Open Chrome and go to chrome://extensions/
• Enable "Developer mode"
• Click "Load unpacked" and select the chrome-extension directory

For detailed setup instructions, see SETUP.md.

📖 Usage

1. Upload Syllabus

Upload a PDF syllabus or type course information. The system will extract:

• Course information (number, name, institution)
• Deadlines (exams, assignments, quizzes)
• Grade breakdown (weightages)
• Topics and required readings

2. Generate Schedule

Connect your Google Calendar and generate a personalized study schedule. The system will:

• Find available time slots ("dead air")
• Create study tasks with specific topics
• Link relevant learning resources (YouTube videos, etc.)
• Insert tasks directly into your calendar

3. Take Practice Tests

Generate practice tests based on your syllabus:

• Select test type (Quiz, Midterm, Final)
• Answer questions based on course topics
• Get instant feedback with explanations

4. Track Mastery

View your 3D concept map:

• See what you know (green) vs. what you haven't touched (red)
• Update mastery levels as you learn
• Track overall progress

5. Active Recall During YouTube

Install the Chrome extension and watch educational videos:

• Extension detects video topics
• Interrupts every 5 minutes with relevant questions
• Reinforces learning through active recall

🎓 Technical Highlights

Privacy-First Approach

• Local LLM Processing: Uses Ollama for syllabus parsing (optional)
• No Data Leakage: Student data never leaves the machine unless explicitly configured
• Secure Storage: Chrome extension uses local storage only

Agentic Capabilities

• Web Automation: Selenium-based course enrollment
• Calendar Intelligence: Finds optimal study times automatically
• Resource Discovery: YouTube, Coursera, EdX integration

Smart Features

• Topic Prioritization: Based on upcoming deadlines
• Adaptive Questions: Generated from actual syllabus content
• 3D Visualization: Interactive Plotly.js concept maps
• Real-Time Updates: Mastery tracking updates visualization instantly

🏆 Demo Highlights for SB Hacks Judges

The "Agentic" Proof

• Show the AI actually clicking a button on Coursera/EdX during demo
• This is the "Aha!" moment that demonstrates true automation

The "Last-Mile" Problem

• Most apps just give you a list of links
• Project Nexus executes the schedule:
  • It puts tasks in your calendar
  • It opens the exact video you need to watch
  • It tracks your progress automatically

Privacy Flex

• Mention that the agent uses Local LLMs (Ollama) to process syllabi
• Student's data never leaves the machine
• Perfect for security-conscious judges

Visual Impact

• Show the transformation: 10-page messy PDF → Living, breathing 3D concept map
• Green/Red nodes show mastery at a glance
• Judges will love seeing the syllabus come alive

📁 Project Structure

backend/
├── context_phase/
│   └── syllabus_parser.py      # PDF/text parsing, deadline extraction
├── agentic_phase/
│   ├── calendar_sync.py        # Google Calendar API integration
│   ├── task_manager.py         # Study task creation and scheduling
│   ├── resource_finder.py      # YouTube, Coursera, EdX resource discovery
│   └── course_enrollment.py    # Selenium-based auto-enrollment
├── active_phase/
│   └── practice_test_generator.py  # LLM-based test generation
├── mastery_phase/
│   └── visualization.py        # 3D concept map generation
└── app.py                      # Flask REST API

frontend/
├── src/
│   ├── components/
│   │   ├── HomePage.js         # Landing page
│   │   ├── SyllabusUpload.js   # PDF/text upload interface
│   │   ├── ScheduleView.js     # Calendar and schedule display
│   │   ├── PracticeTest.js     # Practice test interface
│   │   └── ConceptMap.js       # 3D visualization component
│   └── App.js                  # Main React router
└── package.json

chrome-extension/
├── manifest.json               # Extension configuration
├── content.js                  # YouTube video detection and questions
├── background.js               # Service worker
└── popup.html/js               # Extension popup

🔧 Configuration

Environment Variables (.env)

# Google Calendar API
GOOGLE_CLIENT_ID=your_client_id
GOOGLE_CLIENT_SECRET=your_client_secret
GOOGLE_REDIRECT_URI=http://localhost:5000/auth/callback

# Optional: OpenAI API (if not using Ollama)
OPENAI_API_KEY=your_openai_key

# Ollama Configuration (for local LLM)
OLLAMA_BASE_URL=http://localhost:11434
OLLAMA_MODEL=llama2

# Flask Configuration
FLASK_ENV=development
FLASK_SECRET_KEY=your_secret_key

🐛 Troubleshooting

Chrome Extension not working?

• Ensure extension is enabled in chrome://extensions/
• Check that you're on a YouTube video page
• Verify backend API is running

Selenium/ChromeDriver issues?

• Ensure ChromeDriver version matches Chrome browser
• Install via: brew install chromedriver (macOS) or download from chromedriver.chromium.org

Ollama connection errors?

• Start Ollama: ollama serve
• Verify model: ollama list
• Pull model if needed: ollama pull llama2

See SETUP.md for detailed troubleshooting.

🚧 Future Enhancements

• Mobile app for iOS/Android
• Integration with more learning platforms (Khan Academy, MIT OCW)
• Spaced repetition scheduling
• Collaborative study groups
• AI tutor for personalized explanations
• Analytics dashboard for study habits

📝 License

This project is created for SB Hacks 2024.

👥 Contributors

Created for SB Hacks 2024 at UCSB.

🙏 Acknowledgments

• Ollama for local LLM support
• Google Calendar API for scheduling
• YouTube Search API for resource discovery
• React Three Fiber for 3D visualizations
• Plotly.js for interactive charts

---

Built with ❤️ for students who want to turn chaos into clarity.