Polish — AI-Powered Resume Editor

Polish is a full-stack web application that uses AI to help users write, edit, and improve their resumes. Upload an existing resume or start from scratch, get real-time AI suggestions powered by Gemini, track version history, and export when ready.

Tech Stack

Frontend

• Next.js 14 (App Router), TypeScript, Tailwind CSS v4, shadcn/ui

Backend

• Express.js, TypeScript, Prisma ORM, PostgreSQL, Redis
• Gemini 2.5 Flash for AI suggestions, scoring, and summarization
• JWT authentication (access + refresh tokens)

Infrastructure

• Docker (both services)
• Railway (frontend + backend + PostgreSQL + Redis)
• GitHub Actions CI (type-check on every push)

Local Development

Prerequisites

• Node.js 20+
• Docker (for PostgreSQL + Redis)

Setup

1. Clone the repo

   git clone https://github.com/UNTs-Best/polish.git
   cd polish

2. Start local infrastructure

   docker compose up -d

3. Copy and fill in environment variables

   cp .env.example .env

   Required values:

   DATABASE_URL=postgresql://polish:polish_secret@localhost:5432/polish_db
   REDIS_URL=redis://localhost:6379
   JWT_SECRET=<random 64-char string>
   JWT_REFRESH_SECRET=<random 64-char string>
   GOOGLE_AI_API_KEY=<your Gemini API key>
   PORT=3001
   

4. Run database migrations

   cd server && npx prisma migrate dev

5. Start the backend

   cd server && npm run dev

6. Start the frontend (in a separate terminal)

   npm run dev

7. Open http://localhost:3000

Deployment

Deployed on Railway with four services: frontend, backend, PostgreSQL, and Redis.

Backend environment variables

DATABASE_URL        # from Railway Postgres plugin
REDIS_URL           # from Railway Redis plugin
JWT_SECRET
JWT_REFRESH_SECRET
GOOGLE_AI_API_KEY
NODE_ENV=production
CLIENT_URL          # frontend Railway URL

Frontend environment variables

NEXT_PUBLIC_API_URL  # backend Railway URL
NODE_ENV=production

Branch Strategy

• feature/* → dev (staging)
• dev → main (production)

License

MIT