scame-shield

An AI-powered tool that helps senior citizens instantly detect and avoid online scams with simple, clear guidance

Comment

Every day, thousands of senior citizens in India fall victim to online scams — including fake bank alerts, lottery messages, OTP frauds, and phishing links.

I observed that:

Many seniors struggle to distinguish between genuine and fraudulent messages They often feel uncomfortable using complex technology In critical situations, they don’t know what action to take

This highlighted a deeper issue: 👉 Not just a lack of technology, but a lack of accessible and understandable security tools.

This led to a simple idea:

“What if detecting a scam was as easy as copy-paste?” 🚀 What We Built

ScamShield is an AI-powered web application that allows users to instantly analyze suspicious messages.

Core Functionality: Users paste any SMS, email, or WhatsApp message The system analyzes it using: Keyword detection (e.g., “OTP”, “urgent”, “lottery”) Suspicious language patterns Scam indicators Output: 🚫 SCAM ⚠️ WARNING ✅ SAFE

Along with:

A confidence score Clear, actionable steps (e.g., “Do not click links”, “Contact your bank immediately”) 🧠 How We Built It

The project is designed to be lightweight, fast, and privacy-focused.

Technical Approach: Fully frontend-based architecture No backend, database, or API dependencies All processing happens locally in the browser AI Logic (Explainable Model):

We implemented a rule-based risk scoring system:

𝑅 𝑖 𝑠 𝑘

𝑆 𝑐 𝑜 𝑟

𝑒

𝐾 𝑒 𝑦 𝑤 𝑜 𝑟 𝑑

𝑊 𝑒 𝑖 𝑔 ℎ 𝑡 + 𝑈 𝑟 𝑔 𝑒 𝑛 𝑐 𝑦

𝑆 𝑖 𝑔 𝑛 𝑎 𝑙 𝑠 + 𝐿 𝑖 𝑛 𝑘

𝐷 𝑒 𝑡 𝑒 𝑐 𝑡 𝑖 𝑜 𝑛 Risk Score=Keyword Weight+Urgency Signals+Link Detection Classification: Score ≥ 80 → SCAM Score 50–79 → WARNING Score < 50 → SAFE

This approach ensures:

Fast performance Complete privacy Transparent and explainable results 🎨 Design Decisions

The UI/UX was specifically designed for senior citizens:

Large, readable fonts High contrast color scheme Minimal and intuitive interface No login or signup required Mobile-friendly layout

👉 The goal was to ensure ease of use with zero confusion.

⚔️ Challenges We Faced

  1. Balancing Accuracy and Simplicity

Advanced machine learning models could improve accuracy, but they increase complexity and reduce accessibility.

👉 Solution: We chose a lightweight, rule-based AI system that is fast and reliable.

  1. Designing for Senior Users

Most modern interfaces are not senior-friendly and can feel overwhelming.

👉 Solution:

Simplified UI Clear language Minimal interaction steps

  1. Building Trust in AI

Users may not trust a system if they don’t understand how it works.

👉 Solution:

Provide confidence scores Offer clear explanations Give actionable next steps 📚 What We Learned Simplicity is more powerful than complexity in real-world solutions User experience is as important as functionality AI is only useful if users can understand and trust it Digital safety awareness is a critical need 🔮 Future Improvements Voice input for non-technical users Hindi and regional language support Integration of lightweight ML models (e.g., TensorFlow Lite) Real-time scam database updates WhatsApp chatbot integration ❤️ Impact

ScamShield aims to:

“Empower every senior citizen to stay safe in the digital world.”

Even preventing a single scam can make a meaningful difference.

Built With

Share this project:

Updates