WebTurtle

Tech Stack
Splash screen
Malicious Email Detected!
Parse URL Function

Inspiration

In today's world, social engineering is one of the most prevalent and dangerous cybersecurity threats, exploiting human behavior. Protection against these attacks is vital in ensuring safe online privacy and browsing.

What it does

WebTurtle is a Chrome Extension designed to help detect potential threats on your webpage. Using natural language processing and database analysis, WebTurtle's GoPhish™ engine provides real-time risk assessments. Additionally, WebTurtle includes a toolkit to help dissect and parse hidden actors within your webpage.

How we built it

We split into two teams for frontend and backend development.

Utilizing the Chromium Sandbox, we created JavaScripts to break down the HTML structure and gather information via RegEx. We made reusable components with React to ensure a smooth user experience with robust error handling. Used Supabase to store client information and engagement statistics.

For our logistic regression AI model, we trained it using Python's Scikit-learn pipeline and stored our model in .pkl files, sourcing data from Kaggle. The computation runs on a Python Flask instance deployed via Anaconda and is framed into the extension.

Challenges we ran into

"It's not a Hackathon without AI!" -Sid
Finding an accurate dataset to train our model proved challenging. Many available options were either too small or lacked the diversity needed for our potential use cases.

Accomplishments that we're proud of

We're super proud of our ability to work interdependently and come up with adaptive solutions while delivering a fully functional Chrome extension in under 36 hours.

What we learned

We gained experience building in-house machine learning models, developing React components, configuring DNS, and managing IP network topology. Additionally, we learned how to avoid API rate limits.