AI Menu to CSV Converter

A full-stack React + Express app that uses OpenAI to extract restaurant menu items from images and PDFs, then exports them as CSV.

Quick Start

1. Install dependencies

npm install

2. Configure environment

Copy .env.example to .env and add your OpenAI API key:

cp .env.example .env

Edit .env:

OPENAI_API_KEY=sk-your-actual-openai-api-key
PORT=5000

Alternatively, enter the API key via the Settings page in the UI.

3. Start the backend server

npm run server

Server starts at http://localhost:5000

4. Start the frontend (new terminal)

npm run dev

Frontend starts at http://localhost:5173

5. Or run both together

npm start

---

Environment Variables

Variable	Default	Description
OPENAI_API_KEY	—	Your OpenAI API key (required)
PORT	5000	Backend server port

---

Folder Structure

ai-menu-to-csv/
├── server/
│   ├── index.js          # Express backend
│   ├── uploads/          # Uploaded files (temp)
│   └── data/
│       ├── history.json  # Upload history
│       └── settings.json # App settings
├── src/
│   ├── App.jsx           # React app (all pages)
│   ├── main.jsx          # Entry point
│   ├── styles.css        # All styles
│   └── utils.js          # CSV helpers
├── .env                  # Environment variables
├── .env.example          # Example env file
├── package.json
├── vite.config.js        # Vite + proxy config
├── index.html
└── README.md

---

API Routes

Method	Route	Description
GET	/api/health	Server health check
GET	/api/settings	Get settings
POST	/api/settings	Save settings
GET	/api/history	Upload history
DELETE	/api/history/:id	Delete history entry
GET	/api/stats	Usage statistics
POST	/api/upload	Upload file
POST	/api/process	Process with OpenAI

---

Features

• Drag & drop upload — JPG, PNG, WEBP, PDF
• AI extraction — Uses gpt-4.1-mini via OpenAI Responses API
• PDF support — Direct PDF upload + pdf-parse text fallback
• Editable preview table — Edit/delete rows before export
• CSV export — Download or copy to clipboard
• History — Tracks all uploads with status
• Settings — Configure API key, model, limits
• JSON preview — View raw extracted JSON

---

Troubleshooting

Blank white screen

• Make sure npm run dev is running on port 5173
• Check browser console for errors

ECONNREFUSED / upload errors

• Make sure npm run server is running on port 5000
• The Vite proxy forwards /api/* to http://localhost:5000

OpenAI errors

• Ensure your API key is set in .env or Settings page
• Check you have credits on your OpenAI account
• Model gpt-4.1-mini requires OpenAI API access

PDF not working

• Enable "PDF fallback" in Settings
• Try converting the PDF to an image (JPG/PNG) instead

JSON parsing errors

• The app has robust JSON parsing with multiple fallback strategies
• If a menu is very complex, try a cleaner image/scan

---

Supported File Types

Type	MIME	Notes
JPG/JPEG	image/jpeg	Best results
PNG	image/png	Best results
WEBP	image/webp	Good results
PDF	application/pdf	Falls back to text extraction

Max file size: 10MB (configurable in Settings)