Inspiration

Contractor invoices may exceed agreed project milestone budgets, and manual verification can be time-consuming. We built MilestonePay AI to help businesses identify invoice discrepancies and assess financial risks before approving milestone payments.

What it does

MilestonePay AI is an AI-assisted financial control platform that verifies contractor invoices against agreed milestone amounts. It identifies potential overbilling, generates AI-powered risk assessments, and supports safer payment workflows using PayPal Sandbox.

Key features:

  • Invoice verification: Compare invoice amounts with agreed milestone limits.
  • AI risk assessment: Use Amazon Bedrock Nova 2 Lite to analyze invoice risks and explain potential discrepancies.
  • Payment controls: Flag invoices that require review before payment.
  • PayPal integration: Demonstrate milestone payment workflows using PayPal Sandbox.
  • Audit trail: Review invoice verification and payment activity.

How we built it

We developed the frontend using React and TypeScript, the backend using Java and Spring Boot, and data storage using PostgreSQL. Amazon Bedrock Nova 2 Lite powers AI-assisted invoice risk analysis, while PayPal Sandbox supports payment integration and testing.

Challenges we faced

Integrating AI-generated assessments with backend verification rules and payment workflows was challenging. We also worked on handling invoice discrepancies, validating payment statuses, and connecting the application components reliably.

What we learned

We gained practical experience combining AI with full-stack development and payment processing. We learned that AI can support financial decisions, but backend validation, auditability, and human review remain important safeguards.

What's next

We plan to improve deployment reliability, expand invoice anomaly detection, and strengthen financial reporting and audit capabilities.

Built With

Share this project:

Updates

Submission history