Plainpay

Medical bills explained by AI and paid through PayPal, with a person approving every step that moves money.

Comment

Inspiration

Medical bills shouldn't require a medical degree to understand.

Every day, patients receive complicated medical statements filled with billing codes, insurance adjustments, unfamiliar terminology, and confusing balances. Even after insurance processes a claim, many patients struggle to understand what they actually owe and why.

While working with prior authorizations in a clinic, I witnessed firsthand how confusing healthcare billing can be. Patients aren't always unwilling to pay; sometimes, they simply don't understand what they're being charged for.

That experience inspired Plainpay, an AI-powered medical billing platform designed to make healthcare payments more transparent, understandable, and accessible.

Our mission is simple: Help patients understand their bills, identify potential errors, and pay with confidence.

What It Does

Plainpay transforms complicated medical bills into a clear, patient-friendly payment experience while giving healthcare providers better visibility into their collections.

  1. AI-Powered Medical Bill Explanations

Plainpay uses AI to translate complicated billing terminology into straightforward English. Patients can understand individual charges, insurance adjustments, and their remaining responsibility without needing specialized medical knowledge.

  1. Intelligent Billing Error Detection

The platform highlights potential billing issues, including duplicate charges and suspicious inconsistencies. Instead of automatically changing charges, Plainpay helps patients identify concerns and request a review from their healthcare provider.

  1. Seamless PayPal Payments

Patients can pay their outstanding balance directly through PayPal using a secure checkout experience.

  1. Flexible Monthly Payment Plans

For patients who cannot pay their entire balance immediately, Plainpay offers structured monthly payment options. Each installment is represented by a PayPal invoice, creating a clear payment process.

  1. Centralized Clinic Dashboard

Healthcare providers can manage outstanding balances, monitor payment activity, and review patient billing concerns from one dashboard.

Together, these capabilities make Plainpay more than a payment interface. It becomes a bridge between patients, healthcare providers, and the financial side of healthcare.

How We Built It

Plainpay combines AI, payment infrastructure, and a lightweight web application to create an integrated medical billing experience.

PayPal Integration

We implemented multiple PayPal technologies to support different payment workflows:

  • PayPal Orders v2 API: Processes full-balance payments.
  • PayPal JavaScript SDK: Provides the patient-facing checkout experience.
  • PayPal Invoicing v2 API: Generates invoices for installment-based payment plans.
  • PayPal Webhooks: Communicate payment events to the application and update the clinic dashboard.

Artificial Intelligence

We integrated the Anthropic API to power the patient-facing AI assistant.

Claude generates plain-English explanations of medical bills, answers patient questions, and recommends payment options from a predefined set of permitted plans.

Application Architecture

Our application uses Node.js and Express for the backend, with HTML, CSS, and JavaScript for the frontend.

We integrated AG Grid to build the clinic management dashboard and deployed the application on Render.

The architecture separates AI-generated explanations from financial calculations and payment execution, helping ensure that payment decisions remain controlled by deterministic application logic.

Challenges We Ran Into

1. Making AI Helpful Without Giving It Financial Control

One of our biggest engineering challenges was determining where AI should participate in the billing process.

We wanted Claude to explain charges and suggest payment plans, but we did not want AI to calculate financial obligations independently or initiate transactions.

Our solution was to establish a clear separation of responsibilities.

The application code performs all financial calculations, validates payment amounts, and defines allowable payment plans. Claude only explains the information and recommends options that the application has already approved.

Most importantly, no payment is initiated until the patient explicitly chooses to proceed.

2. Handling Duplicate PayPal Webhook Events

Payment systems must remain reliable even when external services deliver the same event multiple times.

We implemented payment-event handling that prevents duplicate webhook notifications from recording the same payment more than once.

This was an important step toward building a trustworthy payment workflow.

3. Testing Real Payment Integrations

Our initial simulator helped us develop quickly, but testing against PayPal's actual sandbox exposed integration issues that local simulations did not reveal.

We refined the checkout and invoicing workflows based on those findings.

Accomplishments That We're Proud Of

Our biggest accomplishment is delivering a working end-to-end prototype rather than just a concept.

Fully Integrated PayPal Sandbox Payments

Patients can complete full-balance payments through PayPal's sandbox environment.

Working Installment Payment Plans

Patients can select a monthly payment option, with PayPal invoices supporting the installment workflow.

AI-Powered Patient Experience

The application explains complex billing information and helps patients understand their available payment options.

Unified Clinic Dashboard

Healthcare providers can track patient balances, payments, and billing reviews in one place.

37 Automated Tests

We built an automated test suite to validate important application behaviors and improve reliability.

One-Command Offline Demo

The entire application can be demonstrated offline using a single command, making it easier for developers and judges to explore the project.

We're proud that Plainpay demonstrates how AI and PayPal can work together to solve a real-world healthcare challenge.

What We Learned

Building Plainpay reinforced an important engineering principle:

AI should explain complex decisions, while deterministic code should control financial outcomes.

Generative AI is extremely valuable for translating complicated information into understandable language. However, financial calculations, payment authorization, and transaction processing require predictable, validated logic.

We also learned the importance of testing payment integrations against real sandbox environments early in development.

Our local simulator accelerated prototyping, but PayPal's sandbox revealed issues that would otherwise have remained unnoticed.

Finally, we learned that a successful healthcare payment experience is not simply about collecting money. It is about building understanding and trust before asking patients to pay.

What's Next for Plainpay

Our current prototype demonstrates the foundation of a more comprehensive healthcare billing platform.

1. Integration With InsuraMed

We plan to connect Plainpay with InsuraMed's claim data so the platform can generate patient-friendly billing explanations from real claim information.

2. Secure Clinic Authentication and Database

We want to introduce dedicated clinic accounts, authentication, and a persistent database to support more realistic operational workflows.

3. Multilingual Medical Bill Explanations

Healthcare communication should not be limited by language. We plan to enable patients to understand their medical bills in their preferred language.

Our long-term vision is to make medical billing clearer, more accessible, and easier to manage for both patients and healthcare providers.

AI Tools Used

Claude Sonnet 5.5 — Anthropic API

Our primary AI model powers medical bill explanations, patient questions, and payment-plan suggestions.

Claude Haiku 4.5

Serves as a backup model for the AI-powered experience.

Claude for Development

We also used Claude to assist with writing application code, automated tests, and project documentation.

Kokoro Text-to-Speech

We used the open-source Kokoro text-to-speech model to provide voice narration for our demonstration video.

Try It Out

Live Patient Demo:

https://plainpay-qtnz.onrender.com

Clinic Dashboard:

https://plainpay-qtnz.onrender.com/clinic.html

The application is hosted on Render and may take approximately one minute to wake up after inactivity.

Testing a Payment

Open the patient demo, choose the PayPal payment option, and select Pay with Debit or Credit Card in the PayPal checkout window.

Use a PayPal sandbox test card from:

https://developer.paypal.com/tools/sandbox/card-testing/

All transactions are performed in PayPal's sandbox environment. No real money is transferred.

Resetting the Demo

If a sample bill is already marked as paid, open the clinic dashboard and select Reset demo to restore the demonstration data.

Important: All patient information, billing records, and payment amounts used in the demonstration are fictional.

Plainpay — Understand your bill. Trust what you owe. Pay with confidence.

Built With

Share this project:

Updates

Submission history