Inspiration
Healthcare billing can be confusing for patients and time-consuming for providers. Patients often struggle to understand invoices, track payments, and resolve billing issues, while doctors and administrators spend valuable time managing outstanding balances and following up on payments.
We wanted to demonstrate how AI and PayPal could make this process simpler, more transparent, and more secure for everyone involved.
What It Does
Health eRecords AI Billing brings intelligent payment management directly into an existing healthcare platform.
Doctors and patients have dedicated billing dashboards with an AI assistant focused exclusively on financial tasks.
The system is designed to let providers create invoices, identify overdue payments, prepare reminders, and monitor transactions. Patients can review invoices, ask billing-related questions, and complete payments securely through PayPal.
Our central innovation is a three-stage financial workflow:
AI Proposes → Human Approves → PayPal Confirms
AI assists with financial decisions, but users retain control over consequential actions, and payment completion must be verified through PayPal.
How We Built It
The project builds on Health eRecords 2.0, an existing multi-platform application developed with React Native, Expo, TypeScript, and Supabase.
Our planned PayPal integration uses the PayPal Developer Sandbox, Orders API, Invoicing API, and verified webhooks. Supabase manages billing records, notifications, role-based permissions, and audit history.
The existing AI infrastructure provides the foundation for a dedicated Billing AI assistant with restricted tools, human approval requirements, and configurable financial guardrails.
Challenges
One of the biggest architectural challenges is integrating financial operations into a healthcare application without exposing sensitive medical information.
Our approach separates clinical records from billing operations, limits what information can be transmitted to PayPal, and restricts AI actions through server-side permissions.
Another challenge is maintaining consistent payment status across doctor and patient accounts while preventing duplicate transactions and ensuring that notifications reflect verified payment events.
What We Learned
We learned that effective financial AI is about more than automating invoices. It requires trustworthy transaction verification, clear user permissions, human oversight, and a simple experience for both sides of a payment.
What's Next
Our goal is to complete and validate the PayPal Sandbox integration, demonstrate the full doctor-to-patient payment lifecycle, and eventually extend the reusable billing architecture to additional industries supported by Health eRecords.
We envision a payment environment where AI handles the administrative complexity while PayPal provides the trusted transaction infrastructure and users remain in control.
Built With
- cloudflare
- expo.io
- gemini
- react-native
- supabase
- typescript
- vercel
Log in or sign up for Devpost to join the conversation.