Expender
Expense management rebuilt for speed: AI-assisted receipt capture, reliable categorization, and practical reporting for less spreadsheet overhead and cleaner financial workflows.
2025·Finance
Traditional expense tracking breaks down under manual overhead: receipts get lost, categorization drifts, and reporting takes longer than it should. Expender was built to close that gap with practical automation, so teams can move from data entry to financial decisions faster.
What I built:
- Smart Scan - Snap or upload a receipt, AI extracts key info and suggests a category in ~5 secs.
- Multi-currency support across 18+ currencies
- Batch entry for faster input
- Visual analytics with charts and filtering
- CSV export for reporting
- Quota tiers let free users try AI-powered scanning before upgrading
Architecture choices:
- Backend: Python + Boto3 (the AWS SDK for Python)
- Database and Storage: DynamoDB + S3
- AI/ML: AWS Textract for OCR and structured extraction with GPT-4o mini
- Infrastructure as Code: Terraform
- Frontend: React + Next.js with Shadcn/UI
- Authentication: Clerk for user management
- Application Performance Monitoring: Sentry + AWS X-Ray
Impact
Expender moves expense tracking from error-prone work to a repeatable system with AI-assisted intake, structured review, and export-ready reporting. The result is less operational drag and faster, clearer financial visibility.
