What the user first sees when they open the page.
How it works section
Check a payment: works it out for the user, convert, rate checks and checks for scams
Currency exchange rate comparison
How to read a result
Accessiblity

Inspiration

Scams are becoming more common these days, with so many suspicious messages and fake payment requests going around, it's not always easy to tell what's real and what's a scam. For international students and travellers, the friction of transferring money is high, but the anxiety of getting scammed or losing money to hidden exchange rate markups is even higher.

Wise’s mission is all about removing friction and bringing transparency to money. We were inspired to build PayProoof which is a tool to check whether the transaction is a scam or compares rate from different banks within a few seconds. You just need to paste the message or upload the screenshot, and PayProoof will provide clear and instant answer.

What it does

PayProoof is a friction removal tool that checks a payment before you make it. Users can send a message, upload picture and compare currency exchange rates by inserting the values.

The app has four core modes:

Scam Check: Flags high-pressure language, fake landlord requests, and "bank details changed" scams using AI.
Rate Check: Calculates exactly how much money you lose when a provider offers a bad exchange rate.
Convert: Calculates live mid-market conversions.
Auto: Detects what the user is asking for automatically.

How we built it

We built PayProoof as a lightweight, responsive web app:

Frontend: Vanilla HTML, CSS and JavaScript for fast loading, with CSS variables for theme switching and font scaling.
Backend & AI: An LLM connected through a secure backend for OCR and natural language understanding.
APIs: Frankfurter API to fetch European Central Bank reference exchange rates.
Security: A backend proxy keeps our AI API keys hidden from users.

Challenges we ran into

-AI Hallucinations: Since AI isn't always reliable with numbers, we use it to extract structured JSON data, while JavaScript fetches exchange rates and handles calculations for accurate results. -API Failures: We built a fallback system using regex-based local. Analyse to detect common scam phrases and currencies when the AI API is unavailable. -Accessibility: We used CSS clip-path to create distinct shapes and textures for Safe, Caution and Danger alerts, making them easier to recognize without relying on colour alone.

Accomplishments that we're proud of

-Verdict Cards: We designed receipt-style result cards to make financial information easier to read and understand. -Fallback System: Our app can still handle currency conversions and basic scam checks even when the AI API is unavailable. -Accessibility: We added text-to-speech and high-contrast mode to make our website easier for everyone to use.

What we learned

We learned that building reliable AI products means knowing when to use AI and when to rely on traditional code. AI helps us understand user input, while code ensures accurate calculations and reliable results.
Accessibility is a design constraint, we make sure that our website is easy for everyone to use, including people with visual impairments.
We realized that making financial services easier isn't just about saving time. More importantly, it's about helping users feel safe and confident when managing their money.

What's next for PayProoof

Browser Extension: Develop a Chrome/Safari extension that automatically scans websites, such as rental listings, and flags potential scams before users interact with them.
Wise API Integration: Integrate the Wise API to let users make transfers directly through our app after checking exchange rates, with secure payment authorization.
Community Scam Reporting: Build a crowdsourced database where users can report suspicious phone numbers and bank accounts to help others avoid scams.