Inspiration
In many villages, refugee camps, and under-resourced schools, students face three barriers: no internet, no mentors, and no localized learning material. Most ed-tech platforms assume fluent English, strong connectivity, and modern devices. We asked ourselves: “What if a single cheap smartphone could act as a mentor, even offline?” BrainScen was born from that question.
What it does
BrainScen is an AI-powered, offline-first personal mentor for high-school students.
Solves math & science questions step-by-step.
Speaks in local languages (voice & text) with explanations adapted to student level.
Works fully offline with preloaded AI models.
Forms mesh networks between nearby devices to share lessons, quizzes, and progress without internet.
Connects communities by letting teachers/peers contribute notes, voice explainers, and tips.
Gamifies learning with badges and cross-village challenges.
How we built it
Frontend: React + Tailwind + Framer Motion (Apple-inspired clean UI).
Backend: Python (FastAPI/Flask) with SymPy & mathsteps for math solving, Wikipedia/NASA/OpenLibrary APIs for references.
Database & Auth: Firebase for sign-in, progress tracking, and sync.
Offline AI Models: TFLite/ONNX small models preloaded for math/science Q&A.
Mesh Networking: Android layer using Wi-Fi Direct & Bluetooth for device-to-device content sharing.
Design: Inspired by Apple’s UI/UX principles + Dribbble education website references.
Challenges we ran into
Building a mesh-network sync that works seamlessly without internet.
Making AI explanations simple, localized, and level-adaptive.
Ensuring the platform runs on cheap Android devices with low memory.
Designing a premium-feel UI while keeping it lightweight and offline-friendly.
Accomplishments that we're proud of
Created a fully offline AI mentor that actually runs on low-cost phones.
Built a mesh network demo that shares lessons between students without the internet.
Designed an Apple-class UI while keeping accessibility in mind.
Brought together AI + community mentors + gamification into one unique platform.
What we learned
How to combine AI models + symbolic solvers (SymPy) + reference APIs into one education tool.
That offline-first design is the real unlock for under-resourced communities.
Mesh networks are powerful: one connected device can empower an entire village.
The importance of localized, simple explanations over just high-tech features.
What's next for BrainScen
Expand multilingual support with speech-to-text + text-to-speech in local dialects.
Add a teacher dashboard for assigning lessons & monitoring student growth.
Deploy lighter AI models for faster performance on ultra-low devices.
Scale pilot programs in villages and refugee learning centers.
Build a nonprofit + government partnership model to bring BrainScen to millions.
Log in or sign up for Devpost to join the conversation.