Medha.ai — AI-Powered Learning Platform

An intelligent learning companion that transforms education through AI-powered doubt solving, RAG-based study sessions, interactive tutorial support, voice I/O, and personalized analytics. Built as a full-stack application with a modern React frontend, FastAPI backend, and optional browser extension for learning on YouTube.

---

What It Is

Medha.ai is a full-stack, AI-first educational platform that acts as a 24/7 study companion. It combines LLMs (Groq/Gemini), RAG (FAISS + embeddings + Reranking + COT), speech (STT/TTS), and structured learning tools (quizzes, mindmaps, todos) into one cohesive experience—from chatting with AI about your syllabus to watching YouTube tutorials with auto-generated notes and quizzes.

---

What It Does

Area	Capability
Study	AI chat with subject context, persistent sessions, voice input/output
Friend	Mental wellness companion with CBT/mindfulness-style, Hinglish-friendly chat
Sessions	Upload PDFs (syllabus, PYQs, resources), RAG-based Q&A, quiz & mindmap generation
Tutorials	YouTube videos with timestamped notes, AI note prettification, quiz from transcript, mindmaps
Memory Vault	Personal knowledge base: save notes, upload files, semantic search + RAG chat, secure download links
Tasks	Smart to-do with AI suggestions, status tracking, contextual help
Analytics	Single dashboard with chats, sessions, tutorials, quizzes, mindmaps, and TODO progress
Extension	Use Medha.ai on YouTube: notes, AI chat, quizzes, mindmaps without leaving the video

---

Features

AI & Learning

• AI Doubt Assistant — ChatGPT-like study chat with text and voice; subject-specific context
• AI Friend Chat — Empathetic mental-health companion with separate sessions
• RAG-Powered Study Sessions — Upload materials; get answers grounded in your PDFs (FAISS + embeddings)
• Interactive Tutorial Watch — YouTube + AI notes, auto-pause on questions, quiz generation, mind maps
• Memory Vault — Save notes/snippets and upload files; ask questions with RAG, get time-limited download links when needed
• Smart To-Do — Tasks with AI suggestions and step-by-step guidance
• Voice I/O — Speech-to-text (Groq Whisper) and text-to-speech (Kokoro) for hands-free learning
• Learning Analytics — Progress, quiz performance, and study patterns in one dashboard

Platform

• JWT Auth — Register, login, password reset with email OTP
• Multi-Mode Chat — Study mode and Friend mode with history and titles
• Quiz System — MCQs and descriptive questions with AI evaluation and feedback
• Mindmaps — Auto-generated from study sessions and tutorial notes (Graphviz)
• Browser Extension — Notes, AI chat, quizzes, and mindmaps on any YouTube video

---

Benefits

• One place for doubts, tutorials, study materials, tasks, and wellness chat
• Faster revision with RAG (pre-computed embeddings for low-latency answers)
• Better retention via quizzes and mindmaps generated from your content
• Accessible with voice input and TTS for different learning preferences
• Portable with the browser extension for learning while watching videos
• Transparent progress through a single analytics dashboard

---

Screenshots

Screenshots used in this README live in images/.

Landing Page	Study Companion	Tutorial Companion

Organized Study & Notes	Notes	Smart To-Dos

Smart Quizzes	Personalized Quizzes	Browser Extension

---

Architecture & Flow

High-Level System Architecture

flowchart TB
    subgraph Client["Client Layer"]
        Web["React Frontend (Vite)"]
        Ext["Browser Extension"]
    end

    subgraph API["API Layer"]
        FastAPI["FastAPI Backend"]
    end

    subgraph Services["AI & Services"]
        LLM["LLM (Groq/Gemini)"]
        RAG["RAG (FAISS + Embeddings)"]
        STT["Speech-to-Text"]
        TTS["Text-to-Speech"]
        Email["Email (OTP)"]
    end

    subgraph Data["Data Layer"]
        MongoDB[(MongoDB)]
        Azure["Azure Blob (Memory Vault)"]
    end

    Web --> FastAPI
    Ext --> FastAPI
    FastAPI --> LLM
    FastAPI --> RAG
    FastAPI --> STT
    FastAPI --> TTS
    FastAPI --> Email
    FastAPI --> MongoDB
    FastAPI --> Azure

Loading

Request Flow (Authenticated)

sequenceDiagram
    participant U as User
    participant F as Frontend
    participant B as Backend
    participant DB as MongoDB
    participant AI as LLM / RAG

    U->>F: Action (e.g. send message)
    F->>B: HTTP + JWT
    B->>B: Validate JWT
    B->>DB: Read/Write
    B->>AI: Call if needed
    AI-->>B: Response
    B-->>F: JSON
    F-->>U: Update UI

Loading

Project Structure

flowchart LR
    subgraph Root["Medha.ai"]
        FE[Frontend]
        BE[Backend]
        EXT[BrowserExtension]
    end

    FE --> BE
    EXT --> BE

Loading

Tech Stack

Layer	Technologies
Frontend	React 18, TypeScript, Vite, shadcn/ui, Radix UI, Tailwind CSS, Framer Motion, React Router v6, Axios, PDF.js, React Markdown
Backend	FastAPI, Python 3.11, Pydantic v2, MongoDB, JWT auth
AI/ML	Groq API (e.g. Gemma), Gemini, FAISS, SentenceTransformers (all-MiniLM-L6-v2),Langchain, Groq Whisper (STT), Kokoro (TTS)
RAG	FAISS + pre-computed embeddings+Reranking stored in MongoDB
DevOps	Docker, GitHub Actions (e.g. Azure Container Apps), .env configuration

---

How to Run

Prerequisites

• Python 3.11 (backend)
• Node.js 18+ (frontend)
• MongoDB (local or Atlas)
• Graphviz (dot on PATH) for mindmaps
• Optional: Gmail app password for password-reset OTP

1. Backend

cd Backend
cp env.example .env
# Edit .env: GROQ_API_KEY, MONGO_URI, JWT_SECRET_KEY, GMAIL_APP_PASSWORD, etc.

# With uv
uv venv
.venv\Scripts\activate   # Windows
# source .venv/bin/activate  # Linux/macOS
uv pip install -r requirements.txt
uvicorn main:app --host 0.0.0.0 --port 8000

• API docs: http://localhost:8000/docs
• ReDoc: http://localhost:8000/redoc

2. Frontend

cd Frontend
npm install
echo VITE_API_BASE_URL=http://localhost:8000 > .env
# or create .env with: VITE_API_BASE_URL=http://localhost:8000
npm run dev

• App: http://localhost:8080 (or the port Vite prints)

3. Browser Extension (Optional)

cd BrowserExtension
# Update backend/frontend URLs in lib/api.js and popup/popup.js if needed
# Chrome: chrome://extensions → Developer mode → Load unpacked → select BrowserExtension/
# Firefox: about:debugging → Load Temporary Add-on → select manifest.json

• Backend must be running before using the frontend or extension.

---

Environment Variables

Backend (see Backend/env.example)

Variable	Purpose
GROQ_API_KEY	Required for AI (chat, RAG, friend, todo, etc.)
MONGO_URI	MongoDB connection string (include DB name)
JWT_SECRET_KEY	JWT signing (change in production)
GMAIL_APP_PASSWORD	Password reset OTP emails
EMAIL_USER / SMTP_*	Email sender config
AZURE_STORAGE_CONNECTION_STRING	Optional; for Memory Vault

Frontend

Variable	Purpose
VITE_API_BASE_URL	Backend base URL (e.g. http://localhost:8000)

---

Quick Reference

I want to…	Where to look
Run backend	How to Run → Backend
Run frontend	How to Run → Frontend
Run extension	How to Run → Browser Extension
See API endpoints	Backend/README.md + /docs
See frontend structure	Frontend/README.md
Configure env	Environment Variables

---

Project Owner

• Name: Shubham Kumar Gupta
• Role: AI/ML Engineer & Full-Stack Developer
• Email: shubham07kumargupta@gmail.com
• Mobile: +91 8002007238
• LinkedIn: linkedin.com/in/shubhamiitpatna
• GitHub: github.com/SHubhamanjk

---

Medha.ai — full-stack AI for learning: React + FastAPI + LLMs + RAG + voice + analytics.