SHEPHERD - Your AI Digital Safety Companion

AI-powered guardian that detects and prevents digital scams.

Comment

💡 Inspiration

Online scams, fake profiles, and phishing attacks are increasing every day, affecting students, professionals, and even elders.
Most people realize they’ve been scammed only after losing money or data.

We wanted to change that — to build something that not only protects users but also educates them.
That’s how SHEPHERD was born — your AI-powered digital guardian that makes online safety simple, smart, and accessible for everyone.

⚙️ What It Does

SHEPHERD is an intelligent web platform that detects, explains, and prevents online scams in real time.
It scans messages, links, and social profiles using AI-based threat analysis and NLP models to identify phishing, fraud, and misuse.

Unlike traditional tools that only block threats, SHEPHERD teaches users why something is dangerous, building long-term cyber awareness.

🧩 Key Features

  • 🤖 AI Threat Radar – Scans messages and links for phishing or scam patterns.
  • 🧠 Fake Profile & Deepfake Alert – Detects fake accounts and image misuse.
  • 🌍 Cyber Safety Heatmap – Displays scam-prone regions and live cyber threats.
  • 🚨 Guardian Ring SOS – Sends instant emergency alerts to trusted contacts.
  • 🎮 Gamified Cyber Learning – Teaches digital safety through interactive lessons and leaderboards.

🏗️ How We Built It

  • Frontend: React.js and Tailwind CSS for a smooth, responsive user experience.
  • Backend: Flask + Python APIs for secure and fast threat detection.
  • AI Engine: Trained NLP models on phishing, fraud, and scam datasets.
  • Cloud: Firebase + Google Cloud for real-time updates and scalability.
  • Awareness Layer: Gamified learning and safety quizzes for all age groups.

🚧 Challenges We Ran Into

  • Making AI explanations simple enough for non-technical users.
  • Reducing false positives while maintaining fast detection.
  • Designing a trustable UI that feels friendly and not overly technical.
  • Ensuring smooth cross-browser performance for web and extension versions.

🏆 Accomplishments We’re Proud Of

  • Developed a working AI model capable of detecting phishing in text.
  • Designed a human-centric flow that educates while protecting.
  • Built a cross-platform prototype usable by both young and elderly users.
  • Created a concept scalable into a global cyber safety ecosystem.

📚 What We Learned

We learned that cybersecurity is not just about technology — it’s about trust and awareness.
By combining empathy, design, and AI, we discovered how to make digital safety understandable for everyone.

🚀 What’s Next for SHEPHERD

  • 🔊 Voice-based Scam Detection for call and voice message analysis.
  • 🌐 Integration with messaging apps and browsers for real-time protection.
  • 📈 Public Safety Dashboard showing trending scams globally.
  • 🧩 Partnerships with schools, NGOs, and governments to teach cyber awareness.

🧠 Built With

Frontend: React.js, Tailwind CSS
Backend: Flask, Python
AI/ML: scikit-learn, TensorFlow, NLP
Database/Cloud: Firebase, Google Cloud
APIs: Custom Threat Detection APIs

Share this project:

Updates