🌟 Inspiration
India is witnessing an alarming rise in cyber frauds driven by malicious links shared through SMS, WhatsApp, emails, QR codes, and fake websites. A large portion of victims—especially first-time internet users, rural populations, students, and elderly citizens—fall prey due to lack of awareness and delayed detection. Most existing solutions either warn users too late or rely heavily on black-box AI models that are difficult to explain or audit. This inspired us to build Sentinel-X, a transparent, rule-based system that prevents scams before damage occurs and educates users at the moment of risk.
🛡️ What It Does
Sentinel-X is an AI-free, rule-based cyber fraud protection system that:
Blocks phishing, UPI, QR, banking, and scam links before websites load
Analyzes URLs using deterministic rules (patterns, redirects, behavior)
Uses Google Safe Browsing and VirusTotal for reputation verification
Provides explainable alerts showing why a link is unsafe
Opens suspicious links in a secure sandbox mode (no OTP, no payments)
Enables community scam reporting and live threat visibility
Generates cybercrime-ready evidence reports for faster complaints
🛠️ How We Built It
We designed Sentinel-X with a local-first, modular architecture:
Browser Extension intercepts links before page execution
Rule Engine evaluates:
URL patterns and structure
Domain reputation
Redirect chains and behaviors
Signature-based phishing indicators
External intelligence via Google Safe Browsing & VirusTotal
Decision Engine determines Allow / Warn / Block
Sandbox Environment safely inspects suspicious pages
Community Feed aggregates verified scam reports
Logging & Evidence Module stores forensic-ready data
All components are built using proven, lightweight technologies with zero dependency on ML models.
⚠️ Challenges We Ran Into
Designing high-accuracy detection without using ML
Differentiating legitimate websites from highly convincing fake portals
Balancing strict blocking with user convenience
Handling first-time scam links that are not yet blacklisted
Presenting technical security decisions in a user-friendly, explainable way
🏆 Accomplishments That We're Proud Of
Built a fully working end-to-end system, not just a prototype
Strictly complied with non-ML, rule-based requirements
Achieved real-time detection and blocking
Integrated sandbox verification for safe inspection
Implemented community-powered intelligence
Designed a system suitable for low-end devices and mass adoption
📘 What We Learned
Explainability is as important as detection accuracy
Rule-based systems can be highly effective when well-designed
Cybersecurity solutions must focus on user awareness, not just blocking
Community intelligence significantly strengthens fraud detection
Simpler systems are often more scalable and maintainable
🚀 What’s Next for Sentinel-X
QR-code image scanning from posters and printed media
Fake SMS content detection and message-level analysis
Institutional deployment for colleges and offices
Deeper cybercrime portal integration for assisted reporting
National-scale rule updates and threat intelligence sharing
🔑 Final Note
Sentinel-X is not just a security tool — it is a citizen-centric cyber safety platform designed for real-world deployment.
Log in or sign up for Devpost to join the conversation.