"When We Teach, Two Learn"

Known as the Protégé Effect, research shows that teaching others strengthens our own understanding.
Protege flips traditional learning by challenging users to explain concepts to dynamic AI personas: forcing deeper thinking, clearer communication, and true mastery.

• 🧠 Dynamic Persona Teaching: Explain topics to AI personas like a curious child, a seasoned professor, or a historical figure.
• ❓ Adaptive Follow-up Questions: Personas challenge users with real-time, personalized questions to deepen understanding.
• 🎭 Animated Interaction States: Smooth transitions between idle and speaking animations for an immersive experience.
• 📝 Immediate Feedback & Scoring: Receive clarity scores and actionable feedback after each session.
• ⚡ Powered by Gemini Live API: Real-time, streaming interactions that feel fluid, responsive, and human-like.

• Live Demo: https://protege-2025.vercel.app/
• Demo Video: Coming Soon 🎥
• Devpost Entry:

🔨 Built With

1. Clone the repository:

   git clone https://github.com/rappleit/protege.git
   cd protege
   cd protege/live-api-v2-final

2. Install dependencies:

   npm install
   # or
   yarn install

3. Set up environment variables: Create a .env file in the root directory and add necessary GEMINI API KEY

   REACT_APP_GEMINI_API_KEY=<insert API key here>
   

4. Run the application:

   npm run start

   The application should now be running, typically on http://localhost:3000.

• Fawwaz Firdaus (https://www.linkedin.com/in/fawwazfirdaus/)
• Joshua Wee (https://www.linkedin.com/in/joshuawee5/)
• Rachel Lim (https://www.linkedin.com/in/rachellimruien/)
• Chiam Wee Hung (https://www.linkedin.com/in/chiam-wee-hung/)

Protege is an undertaking of the LA Hacks 2025 Hackathon, under the Hack2School Track, the Google Gemini: Chase the Future! challenge, and Major League Hacking Best Use of Gemini API Challenge.