Inspiration
Students having limited, overpriced storage options while other students left apartments open over the summer.
What it does
MadStorage is a peer-to-peer storage marketplace exclusively for UW-Madison students. Students who need summer storage can browse and request available spaces from other students, while hosts can list spare bedrooms, closets, or storage areas at prices our ML model recommends.
How we built it
Data Collection: Built a web scraping pipeline (Playwright + BeautifulSoup) to collect pricing data from 8+ Madison-area storage facilities including U-Haul, Superior Storage, and Fitchburg Self Storage Data Analysis: Ran linear regression on UW enrollment data (R² = 0.916) and commercial pricing to quantify the storage affordability gap Machine Learning: Trained an ensemble model (Random Forest + Gradient Boosting) on scraped rates to power a smart pricing engine that recommends listing prices by neighborhood Full-Stack App: React + TypeScript frontend with FastAPI backend, backed by Supabase for auth, database, and file storage Visualizations: 7 publication-quality charts covering enrollment trends, price regression, demand forecasting, feature importance, and savings projections
Challenges we ran into
We had trouble with integration, git workflow, and implementing features for our database
Accomplishments that we're proud of
We created a full stack website with a server where you can interact with other users within 24 hours.
What we learned
We learned about git workflows, organization, database management, and API endpoints.
What's next for MadStorage
We are looking into how to implementing transportation system for storage, and we are looking to expand to other campuses.
Built With
- css
- fastapi
- matplot
- postgresql
- python
- react
- scikit-learn
- supabase
- typescript
Log in or sign up for Devpost to join the conversation.