Duke Schedule Solver

Find the statistically optimal schedule for YOU (if you're a Duke student)

This is how I got No 8:00 AM's, 2 days off, and all grad reqs filled while graduating a year early.

You can view my writeup here

Uses Binary Integer Programming + CPSAT to maximize of course quality metrics under time/ grad requirement constraints.

Architecture

• Pipeline (scripts/): Ingests catalog + evaluations, applies Bayesian-shrunk quality metrics
• Solver (scripts/solver/): BIP optimizer with configurable constraints
• Backend (backend/): FastAPI API serving the solver
• Frontend (frontend/): React + Zustand wizard UI

Quick Start (Local Dev)

Prerequisites

• Python 3.10+, Node.js 18+
• conda (recommended) or pip

Backend

conda env create -f environment.yml && conda activate solver
pip install -e .  # install project packages for imports
cd backend && python -m uvicorn main:app --reload --port 8000

Frontend

cd frontend && npm install && npm run dev

Open http://localhost:5173

Data Pipeline

python scripts/run_pipeline.py --config config/pipeline_config.json

Pipeline stages:

1. Ingest — Load raw JSON/CSV files
2. Normalize — Parse times, days, course codes
3. Merge — Match evaluations to catalog sections
4. Aggregate — Bayesian shrinkage + z-scores
5. Export — Generate solver-ready JSON

Deployment

See DEPLOY.md for EC2/Docker deployment instructions.

License

MIT — see LICENSE.