Inspiration
Real phishing attacks are so common now that any company can be targeted. We want to use our software to teach them how to protect themselves, to prevent successful phishing attacks in the future.
What it does
We're using deep learning and neural nets to generate completely unique phishing messages that can be sent periodically to users enrolled in training. When a user fails to recognize the message as a phishing attack, a report will be generated so the user can learn from their mistakes and catch the real attacks before phishing becomes a real issue.
How we built it
Our backend is built on python, using a custom code base built on recurrent neural networks. The front end is a web interface built with Django, which also uses python. Using python for both sides made it easier to integrate parts together and collaborate while building the project.
Challenges we ran into
Neural networks are challenging, so it took a lot more planning than coding to get our idea to work as intended.
Accomplishments that we're proud of
Building a fully functional neural network in 36hrs along with a full web interface to enhance usability and support functionality.
What we learned
We all learned a little bit more about Python, and a lot more about neural networks.
What's next for Deep Phishing
We would like to finish building the application and offer it to corporations to train their employees with real world simulation built on a constantly changing base of phishing emails.
Log in or sign up for Devpost to join the conversation.