NeuroStudy-Ai

Login page
Dashboard
Notes Ai Agent
Quiz Ai Agent
Study Plan Ai Agent
Revision Ai Agent

Inspiration

As a student, I constantly struggled with the overwhelming amount of study material. PDFs full of notes took hours to process manually. I wanted a smarter way to study — one that could read a document and instantly generate notes, quizzes, doubt answers, study plans, and revision strategies. That frustration became the foundation of NeuroStudy AI.

What it does

NeuroStudy AI is a multi-agent AI study assistant with 5 specialized agents:

Notes Agent — Extracts key points, definitions, summaries, and topic tags
Quiz Agent — Auto-generates MCQs with difficulty levels and explanations
Doubt Agent — Chat-style Q&A where you ask anything about your document
Planner Agent — Generates a personalized day-by-day study schedule
Revision Agent — Identifies weak topics and creates a focused revision plan

How we built it

Frontend: Next.js 16 + Tailwind CSS, deployed on Vercel
Backend: FastAPI (Python) with 5 independent agent modules, deployed on Render
AI: Groq API using llama-3.3-70b-versatile for fast structured JSON responses
Database: MongoDB Atlas with pymongo for document storage
Auth: Firebase Authentication with Google Sign-In

Challenges we ran into

Groq API key incompatibility with Indian Google accounts — switched from Gemini to Groq
Agents returning empty or malformed JSON — fixed with robust parsing and error surfacing
Firebase auth loop on Vercel — signInWithRedirect failed due to cross-origin cookie restrictions, fixed by switching to signInWithPopup
Vercel build failure due to JSX.Element type not resolving in strict TypeScript — fixed by importing ReactElement directly
MongoDB sync vs async confusion causing silent failures across all agents

Accomplishments that we're proud of

Built and deployed a fully functional multi-agent AI system end-to-end
Designed a premium cyber-tech UI across 5 agent pages with unique per-agent color themes
Each agent produces structured, genuinely useful output — not just raw text
Successfully debugged and resolved 5+ distinct deployment, auth, and API issues
Doubt Agent maintains full conversation history per document for true multi-turn chat

What we learned

Multi-agent architecture requires careful per-agent prompt engineering for reliable structured output
signInWithRedirect is unreliable in deployed environments — always use signInWithPopup
Vercel's production TypeScript build is significantly stricter than local development
Groq LLaMA 3.3 is fast for structured generation but requires explicit max_tokens and robust error handling
Building a real full-stack AI product reveals layers of complexity tutorials never cover