Live demo: https://shooni-frontend.onrender.com
Shooni was originally built in 2022-2023 by a 5-person team through Georgia Tech's CREATE-X program, including Firebase auth, listing search/filter, and a reviews system. The startup didn't continue past that, and the project sat untouched for about three years. In 2026 I revived it as a working portfolio demo: re-provisioned GCP infrastructure from scratch (the original project was gone), reconnected the existing auth and reviews code to it, and implemented favorites and sublease listings — features the original team had planned but never built. All of this is deployed and live at the link above. The original listing scraper (backend/backend_scrape, backend/scraper) no longer works — target sites changed their markup or started blocking the crawler — so the live demo seeds sample listings into Firestore instead (backend/seed_firestore.py, backend/seed_subleases.py) rather than silently faking a working pipeline.
tldr: a website that helps college students find their desired housing much faster and easier
Our objective is to simplify the housing search process by providing a centralized location to help college students find their desired housing. Research into existing solutions from competitors like Apartments.com, Zillow, and Facebook showed that the market lacked a single platform that effectively centralized all desirable aspects, leading to an inefficient and complex search process. To address this, we built a user-friendly forum that lets students filter their search based on preferences, with features like a two-way rating system (agency/homeowner <-> renter), favorites, and sublease listings.
A roommate matching system and a community forum were also part of the original plan; they remain on the roadmap and aren't built yet.
- Frontend: Vue 3 + Vite + TypeScript, deployed as a static site on Render
- Backend: Python, deployed as individual Google Cloud Functions
- Database: Firestore
- Auth: Firebase Authentication
Shooni
├── frontend
│ └── app Vue3 + Vite frontend
│
├── backend Python, deployed as GCP Cloud Functions
│ ├── backend_fetch serves listings from Firestore to the frontend
│ ├── backend_review submit / fetch / delete reviews
│ ├── backend_favorites toggle / fetch favorited listings
│ ├── backend_sublease submit / fetch / delete sublease listings
│ ├── backend_scrape original listing scraper (non-functional, kept for reference)
│ ├── scraper per-site scraper modules used by backend_scrape
│ ├── mock_server local mock API for frontend dev without GCP
│ ├── seed_firestore.py seeds sample apartment listings into Firestore
│ └── seed_subleases.py seeds sample sublease listings into Firestore
│
└── render.yaml Render blueprint for the frontend static site
Frontend
cd frontend/app
npm install
npm run dev
Requires a .env file in frontend/app with Firebase config and backend endpoint URLs (VITE_FIREBASE_*, VITE_*_ENDPOINT_URL, etc.) — see frontend/app/src/firebase.ts and store/index.ts for the full list of variables read at build time.
Backend (without GCP)
backend/mock_server/serve.py serves the same response shape as the real Cloud Functions, using fixture data — useful for frontend development without any GCP setup.
Backend (real deploy)
Each backend_* directory is an independent Cloud Function with its own deploy.sh and requirements.txt. Deploy with gcloud configured for your project:
cd backend/backend_review
./deploy.sh