TAi

Inspiration

Students absorb information at different speeds, and a single lecture often isn't enough for true understanding. We wanted to create a tool that allows students to revisit and interact with course material at their own pace. Our goal was to make learning more engaging, personalized, and accessible.

What it does

TAi lets users upload a PDF of their course content and generates a lecture-style script, which is then read aloud using text-to-speech. While the lecture plays, the main ideas are visually written onto an interactive whiteboard. Users can take notes or solve practice problems directly on the whiteboard, creating a fully immersive learning experience.

How we built it

The backend is powered by Python Flask, while the frontend is built with React. We use Google Text-to-Speech (gTTS) for audio, Gemini for script generation and content analysis, and Excalidraw API to handle the interactive whiteboard. Authentication is handled via Auth0 for a smooth and secure login experience.

Challenges we ran into

One major challenge was animating the whiteboard to simulate real-time handwriting letter-by-letter. Another was seamlessly connecting the flow from LLM-generated text to TTS output to user interaction. Getting all these systems to sync smoothly required careful debugging and creative problem-solving.

Accomplishments that we're proud of

We're proud of delivering a fully functional application that directly addresses the challenge we set out to solve. It blends AI, UI, and interactivity into one cohesive learning tool. Seeing it all come together was incredibly rewarding.

What we learned

We deepened our understanding of LLMs, learned how to build and structure a React frontend, and gained experience in routing data between the backend and frontend. Integrating multiple technologies taught us how to design full-stack applications. We also improved our problem-solving skills through real-time debugging and iteration.

What's next for TAi

Next, we want to make the whiteboard even smarter by detecting and responding to user input in real time. We're also planning to integrate a conversational TTS agent that allows users to ask questions and get spoken feedback instantly. These features will take TAi closer to being a fully interactive, AI-powered tutoring companion.

Built With

• css
• gemini
• google-cloud
• javascript
• llm
• python
• react
• text-extraction
• tts