NJIT Auto Schedule Builder

A Python + Flask service that transforms NJIT's complex course catalog into a clean, accessible API for students to build optimal schedules.

The Challenge

NJIT students faced a frustrating course registration experience:

• Raw JavaScript payloads from the university's system were difficult to parse
• No way to automatically generate conflict-free schedules
• Manual schedule building was time-consuming and error-prone
• Limited filtering options for time preferences and day-of-week constraints

Technical Approach

Backend Architecture

Built a robust Flask service that handles the entire data pipeline:

Data Processing Pipeline:

1. Scraping Layer - Automated extraction of course data from NJIT's catalog
2. Parser Engine - Converts messy JavaScript payloads to clean JSON
3. REST API - Serves structured course and section data
4. Combinatorial Solver - Generates clash-free schedule combinations

Frontend Experience

Created an intuitive interface for schedule building:

Key Features:

• Time-range input controls for availability preferences
• Day-of-week filters for flexible scheduling
• Sequential schedule preview with conflict detection
• Clean, mobile-responsive design

Algorithm Design

The core scheduling algorithm uses constraint satisfaction:

def generate_schedules(courses, constraints):
    """
    Generate all possible conflict-free schedules
    given course requirements and user constraints
    """
    valid_schedules = []
    
    for combination in itertools.product(*course_sections):
        if not has_time_conflicts(combination):
            if meets_constraints(combination, constraints):
                valid_schedules.append(combination)
    
    return sorted(valid_schedules, key=schedule_score)

Impact & Results

Student Experience:

• Reduced schedule planning time from hours to minutes
• Eliminated manual conflict checking
• Provided multiple optimal schedule options

Technical Achievements:

• 100% uptime during peak registration periods
• Sub-second response times for schedule generation
• Clean API design adopted by other student developers

What I Learned

This project taught me valuable lessons about:

• Data Pipeline Design - Building robust systems that handle messy external data
• Algorithm Optimization - Balancing completeness with performance in combinatorial problems
• User-Centered Design - Creating interfaces that solve real student pain points
• API Design - Building clean, documented endpoints that others can build upon

Next Steps

If I were to continue this project, I would:

• Add GraphQL endpoint for more flexible data querying
• Implement caching layer for improved performance
• Build mobile app with push notifications for seat availability
• Add machine learning for schedule preference learning
• Create admin dashboard for monitoring and analytics

This project demonstrates my ability to build full-stack solutions that solve real-world problems, from data processing to user experience.