Inspiration
Choosing courses and mapping them to future careers is overwhelming for students. Academic advisors are stretched thin, and students often lack clear guidance. We wanted to build an AI-driven copilot that personalizes academic decisions and connects them to career goals.
What it does
Career Path Finder is an AI-powered student success platform that:
Predicts a student’s likelihood of success in a given course.
Recommends optimal next-term courses (with difficulty labels, prerequisites, and probability of success).
Highlights popular courses and peers with similar paths.
Suggests career paths aligned with the student’s academic profile.
How we built it
Backend: FastAPI service with endpoints for prediction, recommendation, and career insights.
Data Layer: Neo4j graph database modeling students, courses, degrees, and terms.
ML Model: PyTorch neural network trained on student performance data, wrapped with joblib scaler + feature encoding.
Frontend: Next.js dashboard for students to log in, view predictions, recommendations, and track their journey.
Challenges we ran into
Handling students with little or no course history (nulls/empty values in Neo4j).
Keeping the model predictions consistent when categorical features differed across students.
Managing multiple lockfiles and Next.js root conflicts when wiring up the frontend.
Debugging Neo4j queries for degree-specific vs cross-department course recommendations.
Accomplishments that we're proud of
Built a full stack system that connects graph data + ML + web frontend in under one project.
Achieved interpretable predictions with clear success labels (“High”, “Medium”, “Low”).
Designed APIs that are clean, reusable, and easy to extend for advisors or admins.
Created a working prototype where a student can log in and immediately see personalized career guidance.
What we learned
The importance of defensive coding when student data is incomplete.
How graph queries can power personalized recommendations at scale.
Practical challenges of connecting a trained ML model into a live API.
How to make predictions interpretable and student-friendly (not just raw probabilities).
What's next for Career Path Finder
Integrate real student datasets and validate predictions against actual outcomes.
Add a career recommendation engine that maps course histories to job titles/skills.
Build advisor dashboards with cohort analytics.
Enhance the frontend with progress visualizations and what-if simulations (“If I take these courses, how does it affect my graduation date?”).
Explore integrating LLMs for natural-language career advice.
Log in or sign up for Devpost to join the conversation.