Homepage

ScamSniff AI – Hackathon Story

Inspiration

Online fraud is evolving faster than most people can detect it. From phishing messages to fake websites and manipulated screenshots, I’ve seen how easily users—especially students and small business owners—fall victim to scams. This inspired me to build a tool that can act as a first line of defense before damage happens.

What I Built

I developed an AI-powered scam detection platform that analyzes:

Messages (SMS, WhatsApp, email)
Links and URLs
Screenshots and images

The system outputs a risk score and a short explanation of why something may be a scam, making it easy for non-technical users to understand.

Mathematically, the risk model can be simplified as:

[ R = \sigma(w_1 f_{text} + w_2 f_{url} + w_3 f_{image}) ]

where (R) is the final scam probability score, and (f) represents extracted features from different inputs.

How I Built It

I built the project using:

A web-based frontend for user interaction
An AI analysis layer for text and image classification
API integration for real-time scanning and response generation
A dashboard to visualize threat patterns globally

The focus was on speed, simplicity, and explainability so users don’t just get a “yes/no” answer but understand the reasoning.

What I Learned

This project taught me:

How to combine multiple AI models into one system
The importance of explainable AI in security tools
How user experience is just as important as model accuracy
How to design systems that work under real-world uncertainty

Challenges Faced

One of the biggest challenges was reducing false positives while still catching subtle scams. Another was handling different types of inputs (text, URLs, images) in a unified pipeline.

I also struggled with balancing model complexity and response speed—because security tools must respond instantly to be useful.

Conclusion

This project strengthened my belief that AI can be used not just for automation, but for digital protection and trust-building online. My goal is to keep improving it into a real-world fraud prevention system that can help everyday users stay safe.