Inspiration

We were inspired by the massive annual losses caused by cyber fraud. Most tools only help after a loss occurs, so we built a platform to educate users before they are scammed.

What it does

ScamGuard is an AI-powered awareness system that detects scams in real-time using Google Gemini. It offers risk scoring for messages, a community-driven contact verification database, and interactive quizzes.

How we built it

We used a four-layer stack: Python and Flask for the backend, Supabase (PostgreSQL) for the database, and Bootstrap 5 for the frontend. AI detection is powered by the Google Gemini API.

Challenges we ran into

Integrating real-time AI risk scoring with explainable results was a key challenge. We also focused on creating a unified dashboard that tracks both financial impact and scam types accurately.

Accomplishments that we're proud of

We successfully built a platform that doesn't just give a risk score but explains why content is a scam. We also developed a structured 60-question learning system to reinforce proactive defense.

What we learned

We learned how to leverage LLMs for cybersecurity awareness and the importance of community-driven data in improving detection accuracy.

What's next for ScamGuard

Our roadmap includes developing native iOS/Android apps, browser extensions, and voice-based scam detection to protect users across all communication channels

Built With

bootstrap
css3
flask
google-gemini-api
html5
javascript
postgresql
python
supabase

Submitted to

LUMA Pitch Competition (July 3rd - 10th)

Created by

Daniyal Arain (24SW039) — Group Leader & Backend Engineer

As the group leader and primary backend developer, Daniyal served as the backbone of the entire ScamGuard project. He architected and implemented the complete backend system using Flask and Python, designed the database schema, and ensured seamless integration across all system components. He successfully integrated the Gemini AI API, bringing intelligent scam detection capabilities to life. Beyond the technical work, he conducted thorough end-to-end testing to validate system reliability and personally deployed the entire application on Render, making ScamGuard publicly accessible. One of his most significant contributions came over a single weekend, during which he single-handedly developed the core logic and main functionalities of the system — a remarkable effort that formed the foundation everything else was built upon. As group leader, he also mentored and guided all team members throughout the project, ensuring every person had a clear understanding of their responsibilities and that the team moved forward cohesively.

Daniyal Arain -24sw039
Software Engineering Student
Sami Memon (24SW127) — Video Content Creator.

I was responsible for producing all the video content featured in ScamGuard. I created 10 original educational videos, each one focused on a specific type of scam covered in the platform, including phishing, cryptocurrency fraud, identity theft, tech support scams, deepfakes, and more. Each video was designed to break down complex scam tactics into simple, engaging, and easy to understand visual stories. My videos serve as the first learning touchpoint for users exploring each scam category, adding a multimedia dimension to the platform that makes the educational experience more impactful and memorable.

Abdul Sami Memon
Fatima (24SW021) — Content & Data Specialist I was responsible for all the research, data collection, and content creation that powers ScamGuard's educational core. I gathered and organized all the information related to the 10 scam types covered in the platform, ensuring the content was accurate, relevant, and easy to understand. I also developed the complete question bank for the quiz system — including 100 practice questions and 60 main quiz questions across three difficulty levels (Easy, Medium, and Difficult) — along with the detailed explanations and answer feedback for each question. My work forms the knowledge foundation of ScamGuard, directly helping users recognize and protect themselves from real-world scams.

fatima24sw021
Tariq Jameel (24SW069) — Frontend Developer & UI Designer I was responsible for the complete frontend development and UI design of ScamGuard. I designed and implemented the entire user interface using Bootstrap 5 and custom CSS, focusing on creating a modern, clean, and responsive dark-themed experience. I made sure the platform works seamlessly across all screen sizes — from mobile phones to large desktop monitors. Every visual element the user interacts with — the layouts, buttons, cards, navigation sidebar, and page transitions — was designed and built by me with the goal of making ScamGuard feel polished, professional, and easy to use.

Tariq Jameel