🍽️ SplitBite

Smart receipt splitting that actually works. Upload a receipt photo, automatically extract items, and split costs instantly with friends.

🚀 Live Demo

What I Built

I developed SplitBite as a full-stack solution to solve the annoying problem of splitting restaurant bills. Instead of manually calculating who ordered what, users simply snap a photo and get instant, accurate item extraction and cost splitting.

Key Features:

• 📱 Automated Receipt Scanning - OCR technology for accurate item extraction
• 👥 Effortless Bill Splitting - Assign items to people, auto-calculate totals with tax/tip
• 🔐 Secure User System - JWT authentication with bcrypt password hashing
• 📊 Receipt History - Track spending patterns and restaurant visits
• ⚡ Real-time Processing - Instant receipt parsing and split calculations

Tech Stack

Frontend: TypeScript, HTML5, CSS3 (responsive design) Backend: Node.js, TypeScript, Express.js Database: MongoDB with Mongoose ODM Processing: AWS Textract (OCR), OpenAI (text parsing) Storage: AWS S3 for receipt images Auth: JWT tokens with secure session management

Architecture Highlights

• Service-oriented design with separate modules for S3, OCR, text parsing, and split calculations
• Robust error handling with fallback parsing when cloud services fail
• RESTful API design with proper HTTP status codes and input validation
• Environment-based configuration for seamless dev/prod deployment
• Scalable file storage with signed URLs for secure image access

Quick Start

git clone https://github.com/znimer-dev/splitbite.git
cd splitbite/backend
npm install
npm run dev
# Visit http://localhost:5000

Key Implementation Details

Receipt Processing Pipeline:

1. AWS Textract extracts raw text from receipt images
2. OpenAI parses text into structured data (restaurant, items, prices)
3. Custom algorithms handle tax/tip distribution and splitting logic
4. Fallback to regex-based parsing for reliability

Database Design:

• User authentication with salted password hashing
• Receipt documents with embedded item arrays and split calculations
• Restaurant history tracking for spending analytics
• Database indexing for optimized queries

Security & Infrastructure:

• File upload restrictions (type and size validation)
• Secure environment variable management
• CORS configuration for cross-origin requests
• Input validation on all API endpoints

Why I Built This

Splitting bills manually is tedious and error-prone. I wanted to create something that's seamless, snap a photo, get accurate results instantly. Not only that, but this is my first true full stack project, a big challenge was learning as I was going, and debugging problems I've never encountered before.

What's Next

• Mobile app development (React Native)
• Integration with payment platforms (Venmo, PayPal)
• More Splitting Options
• Group expense tracking and analytics
• Multi-language receipt support
• Enhanced OCR accuracy improvements

---

Built by: Zaid Nimer

Contact: [[email protected]]

Portfolio: [https://www.linkedin.com/in/zaid-nimer/]