Inspiration
Inspired by CougarGrades, we set out to solve a critical challenge faced by UH students: creating an optimal course schedule that balances workload while ensuring on-time graduation. Students often struggle to navigate complex degree requirements, prerequisite chains, and credit hour balancing when planning their academic journey. We wanted to build a tool that makes degree planning as simple as filling out a form and clicking a button.
What it does
Cougar Degree is a full-stack intelligent degree planning application that revolutionizes academic scheduling for UH students. Users simply select their major, starting semester, and any completed coursework through a clean web interface. Our backend powered by Google Gemini AI processes this information using structured degree requirement data extracted from official UH degree checklists. The system generates a personalized 8-semester degree plan that enforces prerequisite requirements, balances credit loads around 15 hours per semester, and ensures on-time graduation. The plan displays semester-by-semester course schedules with credit breakdowns, prerequisite tracking, and progress visualization, giving students a clear, actionable roadmap to their degree.
How we built it
We built a full-stack application with a lightweight vanilla JavaScript frontend styled with custom CSS featuring UH branding colors. The backend is a Node.js Express REST API that implements intelligent course scheduling algorithms including topological sorting for prerequisite resolution and greedy bin-packing for credit hour balancing. We integrated Google Gemini 2.5 Flash API as our AI engine to generate optimized degree plans. The system uses structured JSON files containing course catalogs, prerequisites, and degree requirements manually extracted from official UH degree checklists. We implemented security best practices including Helmet middleware, CORS whitelisting, and rate limiting. The architecture features a hybrid approach: AI-powered plan generation with deterministic validation fallback to ensure reliability even if the AI service is unavailable.
Challenges we ran into
Building a robust degree planner presented several technical challenges. Implementing the prerequisite resolution algorithm required careful graph theory application to detect circular dependencies and ensure valid course sequencing. Balancing credit hours across semesters while respecting prerequisites proved complex, we had to develop a constraint satisfaction system that treats degree planning as an optimization problem. Extracting and structuring UH degree requirements from PDF checklists into machine-readable JSON format was time-intensive and error-prone. Integrating the Google Gemini API required extensive prompt engineering to ensure consistent, valid JSON responses. We also had to design a graceful fallback system for when the AI service is unavailable, implementing template-based plan generation as a backup. Managing the frontend-backend communication and handling edge cases like transfer credits and course substitutions added additional complexity.
Accomplishments that we're proud of
We successfully built a functional, end-to-end degree planning system in a hackathon timeframe. We're proud of implementing sophisticated algorithms for prerequisite resolution and credit balancing that actually work correctly. Creating structured JSON datasets for five different majors from UH's degree checklists was a substantial data engineering achievement. Our hybrid AI architecture combining Gemini's intelligence with deterministic validation ensures both smart planning and reliability. The system generates valid, realistic degree plans that students could actually follow. We're particularly proud of the clean, responsive frontend that makes complex degree planning feel simple and the robust API design with security best practices built in from day one.
What we learned
We learned that building an AI-powered application requires balancing intelligent automation with deterministic validation, you can't rely solely on AI for critical logic. Working with graph algorithms taught us how academic concepts like topological sorting solve real-world problems. We discovered that data engineering (extracting and structuring UH degree requirements) takes significantly more time than anticipated but is crucial for accuracy. Prompt engineering for consistent AI output is an iterative process requiring careful testing and refinement. We learned the importance of building fallback systems and graceful degradation into web services from the start. The hackathon environment taught us to prioritize core functionality over feature creep and to make quick architectural decisions with limited information. Team coordination and clear communication about who's working on what proved essential for avoiding merge conflicts and duplicate work.
What's next for Cougar Degree
Our immediate roadmap focuses on completing the data engineering work to support all UH majors beyond our current five, expanding to all 100+ degree programs across the university. We plan to integrate MongoDB for user account creation and plan persistence, allowing students to save, edit, and share their degree plans. Incorporating CougarGrades.io historical grade distribution data will enable difficulty-aware course recommendations. We want to add summer session support for accelerated graduation paths and integrate real-time course availability data from UH's registration system. Future features include RateMyProfessor integration for instructor recommendations, visual prerequisite tree diagrams, GPA projection tools, minor and certificate planning, transfer credit evaluation, and mobile apps for iOS and Android. Long-term, we envision building an analytics dashboard showing popular course sequences and graduation timeline trends to help UH continuously improve degree program design.
Log in or sign up for Devpost to join the conversation.