End‑to‑end demo implementation (NestJS + React + TailwindCSS)
Live frontend: https://pery-frontend.vercel.app/
Warm‑up backend (first request if idle): https://pery-backend.onrender.com/introduction/cat
The backend is hosted on Render’s free tier. If no traffic is received for ~15 minutes it spins down; the first request will take ~30 s while the instance wakes up.
/ ← this README (root‑level overview)
├─ backend/ ← NestJS API → README explains endpoints & running
└─ frontend/ ← React UI → README explains dev / build / deploy
Each sub‑folder has its own README with detailed instructions.
# clone
git clone https://github.com/ary428/pery.git
cd pery
# install all workspace deps
npm install
# start backend & frontend concurrently (default ports 4000 / 5173)
# ↳ backend: http://localhost:4000
# ↳ frontend: http://localhost:5173
npm run devBoth apps share zero configuration: the frontend auto‑detects whether it’s running from localhost and calls the proper backend URL.
| Service | URL |
|---|---|
| Frontend | https://pery-frontend.vercel.app/ |
| Backend | https://pery-backend.onrender.com |
Tip: if the backend was asleep, open the warm‑up link above or retry after a few seconds.
This project implements all requirements from the Pery Full‑Stack assignment:
- GET
/introduction/:articleName– returns first paragraph + caching 5 min - POST
/user– registers user, stores preferred language, returns token - Language resolution: token →
Accept‑Language→ defaulten - Supports
en,fr,nl(extensible)
- Sign‑up → fetch article flow following Figma styling
- Responsive (desktop / tablet / mobile), TailwindCSS
- Loading spinner & fade/slide page transitions
- FAQ button, “Start Over” navigation
- Backend on Render, Frontend on Vercel
- Public repo (this one) with clear docs (you are here!)
- Replace in‑memory cache with Redis
- Persist users in DB
- Extend language list dynamically via
/meta/languages - Add e2e tests (Playwright) & CI pipeline
© 2025 Pery Demo