Mentor AI

Command Center dashboard — track your learning progress, current streak, and quick access to every feature at a glance.
Ask MentorAI any doubt and get clear, step-by-step explanations powered by the Gemini API in real time.
AI-generated personalized roadmap — enter any goal and get a week-by-week learning path with topics and time estimates.
Quiz Studio — turn your notes into an instant quiz to test what you've learned and reinforce your knowledge.

Inspiration

Most students hit a wall when they're stuck on a doubt at 11 PM with no one to ask, or they want to learn a new skill but don't know where to start — which topics matter, in what order, and how long it'll realistically take. We wanted to build something that acts like a personal mentor: available anytime, breaks down any doubt into simple explanations, and turns a vague goal like "learn Data Analytics" into an actual structured plan.

What it does

MentorAI is an AI-powered learning companion built around three core features:

Learning Map — Enter any goal (like "Learn DSA" or "React advanced concepts"), pick your level and timeline, and Gemini generates a personalized, week-by-week roadmap with topics, subtopics, resource suggestions, and an estimated difficulty/pace assessment.
AI Mentor (Doubt Solver) — A real-time chat where students can ask any doubt — code, concepts, anything — and get clear, step-by-step explanations rendered with proper markdown formatting (code blocks, lists, headings). Replies can also be read aloud with one tap.
Quiz Studio & Progress Tracking — Students can mark topics as done, track their focus streak, and reinforce what they've learned.

Everything lives inside a clean, focused dashboard so students always know what to do next.

How we built it

Frontend: React + Vite, Tailwind CSS for styling, custom components for the roadmap timeline, chat interface, and progress tracking, deployed on Netlify.
Backend: FastAPI (Python) serving REST endpoints, with SQLite for storing users, roadmaps, and progress. Deployed on Render.
AI: Google's Gemini API powers both roadmap generation (structured JSON output) and the doubt-solving chat, with conversation history passed in for context-aware follow-ups.
Auth: Simple email/password authentication with hashed passwords, scoped per user so roadmaps and progress persist across sessions.

Challenges we ran into

Getting the AI to reliably return structured, parseable roadmap data (topics, subtopics, realistic time estimates) instead of free-form text required careful prompt design.
Balancing an ambitious visual design (glassmorphism, glowing gradients, smooth animations) with performance and readability.
Debugging deployment-specific issues — CORS between the Netlify frontend and Render backend, and making sure environment variables and build paths were correctly configured across two separate hosting platforms.

Accomplishments that we're proud of

A fully working end-to-end product — auth, AI roadmap generation, AI doubt chat with text-to-speech, and progress tracking — built and deployed within the hackathon timeline.
A cohesive, polished UI that doesn't feel like a typical hackathon prototype.

What we learned

How to design prompts that get consistent, structured output from an LLM instead of unpredictable free text.
Practical experience deploying a decoupled frontend/backend architecture across two different platforms and debugging real production issues like CORS.

What's next for MentorAI

Voice input for the doubt chat (speech-to-text), so students can ask questions hands-free.
Adaptive quizzes generated directly from a student's own notes.
Migrating from SQLite to a persistent hosted database for reliability at scale.