Inspiration
We were inspired by the real-life stories of our families and communities. The rise of "Hi Mom" scams, SIM swap fraud, and phishing attacks targeting mobile money users is not just a news headline—it's a constant threat that erodes trust in the digital financial systems that empower millions. We built Guard to be the shield that protects the hard-earned money of everyday people.
What it does
Guard is an AI-powered security layer that integrates directly with mobile money platforms. It detects suspicious transactions in real-time (like a large transfer to a new number) and instantly sends a clear SMS alert to the user, asking them to approve or block the payment with a single reply. It puts the power to stop fraud directly into the user's hands, using the simple SMS technology available on every phone.
How we built it
Backend & AI Logic: Built with Node.js and Express, it handles the core decision-making and risk analysis.
SMS Communication: Integrated the Twilio API to send and receive text messages instantly and reliably.
Data Storage: Used MongoDB Atlas, a cloud database, to securely store and manage transaction data and user interactions in real-time.
Dashboard: A live-updating web dashboard built with HTML, CSS, and JavaScript provides a visual overview of fraud activity.
Prototyping: Used ngrok to create a secure tunnel for testing and Figma to design the user flow.
Challenges we ran into
Our biggest challenge was creating a seamless, real-time user experience by orchestrating multiple services (Twilio, our server, the database) to work together flawlessly. Debugging this flow required meticulous monitoring of several logs simultaneously. We also focused intensely on designing for accessibility, ensuring our solution worked perfectly on the most basic feature phones, not just smartphones.
Accomplishments that we're proud of
We are incredibly proud of building a functional, interactive prototype that demonstrates the entire user journey from threat to solution in real-time during a demo. We created a system that is both technically sophisticated and incredibly simple for the end-user. Most of all, we're proud of designing a solution that has genuine potential to make a tangible, positive impact on financial security in our communities.
What we learned
We learned that the most impactful technology is often invisible. This project taught us the power of leveraging universal platforms like SMS to create inclusive solutions. We gained hands-on experience in full-stack integration, connecting cloud databases, third-party APIs, and front-end visualization into a single, cohesive application. Beyond code, we learned to prioritize user experience and accessibility above technical complexity.
What's next for AI-Powered Fraud & Scam Detection for Mobile Money
The future for Guard is focused on making the AI smarter and the platform stronger. Our next steps include:
Advanced Machine Learning: Training our model on real, anonymized transaction data to better identify sophisticated fraud patterns and reduce false positives.
Provider Partnerships: Piloting the technology with a mobile network operator to integrate directly into their USSD flow.
Community Reporting Network: Enhancing the reporting system to create a crowd-sourced blacklist of fraudulent numbers, making the entire ecosystem safer for everyone.
Multi-Platform Expansion: Extending alert channels to include popular platforms like WhatsApp and Telegram for users with data access.
Log in or sign up for Devpost to join the conversation.