mirage

• 

  home page

• 

  log in

• 

  facial recognition setup

• 

  file manager

• 

  intruder alert

Inspiration

Caden's Instagram account got hacked, and that made me realize how exposed personal files and accounts can be. That experience inspired us to build Mirage, a system that not only protects sensitive documents but also turns unauthorized access into something users can investigate.

What it does

Mirage is a secure file explorer that protects sensitive files with password login plus facial recognition. If an attacker gains access, Mirage serves decoy documents instead of the real ones while logging evidence and activity to help users understand and respond to the breach.

How we built it

We built Mirage with React, Node.js, and Express, using SQLite for storage and face-api.js for facial recognition. We used Ollama with Gemma to generate decoy documents and Vultr to support deployment.

Challenges we ran into

Our biggest challenges were integrating facial verification smoothly into login, designing a clean and believable UI, and coordinating changes across a team without stepping on each other’s code.

Accomplishments that we're proud of

We’re proud of creating a polished interface and a security workflow that feels both unique and practical. We’re also proud that our facial detection achieved over 90% accuracy in testing.

What we learned

We learned that working in a team requires constant communication or else your code gets merge conflicts on GitHub.

What's next for mirage

We want to deploy Mirage for real-world use, improve the facial recognition pipeline, and explore stronger AI models for more convincing decoy document generation.

Built With

• express.js
• gemma
• node.js
• ollama
• react
• sqlite
• vultr