ByteWave

🌊 ByteWave: AI-Powered Physics Mastery The Problem Students often re-read textbooks blindly or watch generic YouTube videos without realizing what specific concepts they actually misunderstand. They lack a feedback loop to pinpoint their exact knowledge gaps, making standard studying highly inefficient. This inspired us on making a study platform that can help student study more efficiently.

The Skill Map: The platform features 10 core physics topics arranged in a visual constellation. Mastered nodes glow vibrantly, while unstarted concepts remain dim, making progress obvious at a glance. Interactive Cases: Instead of multiple-choice tests, students pick from 50+ real-world physics scenarios (e.g., ramps, collisions, graphs) that force them to conceptually apply their knowledge. AI Gap Analysis: Students explain their reasoning. The core engine analyzes every sentence to pinpoint exactly where their mental model broke down, providing highly specific, encouraging feedback rather than just a pass/fail grade. Adaptive Mastery: The dashboard instantly updates after each case, sorting topics into "Next for you" or "Ready to master" so students always know exactly what to study next. Instant Animation Sandbox: A built-in AI Chatbot lets students ask ad-hoc physics questions. The AI instantly generates a beautiful, pre-rendered educational animation to help visualize the concept. Technology Stack ByteWave is a fast, modern web application powered by seamlessly integrated AI:

Frontend: Built with React and Vite for a lightning-fast user experience. Features a custom dark mode, glassmorphism UI, and interactive SVG graphics. Deployed on Vercel. Backend: Powered by Python and FastAPI to handle data and AI requests quickly. Database: Uses PostgreSQL (hosted on AWS RDS) to securely save and track student progress over time. Deployed via Render. AI & Intelligence: Uses the MiniMax-M2.5 language model (routed via OpenRouter) to naturally converse with students and assess their physics answers. Features custom text-classification to instantly match student questions with high-quality, Python-generated visual animations.

During this project, our biggest challenges is on the Manim code generator especially on making the animation. This requires 2 API calls which translate user request to physics, then use that to write the Manim. We are able to fix this challenge however, and is able to create a custom video for user requests.

Built With

• css
• fastapi
• html5
• javascript
• minimax
• postgresql
• python
• react
• vite