Engineering Tutor AI: Agentic Research Assistant for Engineering Education

A smart, interactive platform to revolutionize engineering education using Retrieval-Augmented Generation (RAG), AI voice tools, and visual learning enhancements.

---

🎮 Demo Videos

• 🧪 Product Walkthrough
• 🛠️ Code Explanation

---

🧠 Key Features

1. 🗣️ Conversational Podcast Generator

• Converts technical documents into dialogue scripts.
• Uses TTS to generate MP3 audio.
• FFmpeg + eSpeak for high-quality synthesis.

Requirements:

• FFmpeg
• eSpeak

Packages:

npm install @ffmpeg/ffmpeg @ffmpeg/core @ffmpeg-installer/ffmpeg fluent-ffmpeg say

2. 🧠 Interactive Mind Map Generator

• Creates readable mind maps with Dagre layout.
• Interactive fullscreen view.

Packages:

npm install dagre reactflow react-icons react-tiny-popover

3. 📂 Multi-File Upload Support

• Upload PDFs, DOCX, PPTX, TXT with real-time progress.
• Action menu: rename, delete, convert to podcast, generate mind map.

4. 🔗 Chain-of-Thought Reasoning with RAG

• Answers pulled directly from uploaded docs.
• Uses FAISS for vector search.

Packages:

npm install @langchain/community @langchain/core @langchain/google-genai faiss-node

5. 💾 Persistent Chat History

• MongoDB for storing user sessions.
• Load, delete, manage chats via modal interface.

6. 🎤 STT & TTS Interaction

• STT: Voice-to-text using browser APIs.
• TTS: Replies spoken using real-time engines.

Packages:

npm install @google-cloud/text-to-speech say

7. 🔍 Deep Search Integration

• Uses DuckDuckGo API for advanced search.
• Decomposes queries, aggregates and caches results.

Packages:

npm install duck-duck-scrape axios node-cache

---

⚙️ Setup & Installation Guide

Prerequisites

• Node.js (v18+)
• MongoDB
• FFmpeg
• eSpeak

Installation Steps

1. Clone the Repository

git clone -b Team-4 https://github.com/spkkarri/iMentor.git

2. Install System Dependencies

FFmpeg:

• Windows: FFmpeg for Windows
• macOS: brew install ffmpeg
• Linux: sudo apt-get install ffmpeg

eSpeak:

• Windows: eSpeak Windows
• macOS: brew install espeak
• Linux: sudo apt-get install espeak

3. Terminal 1: Backend Setup

cd server
npm install

Create .env file in /server:

PORT=5007
MONGO_URI=mongodb://localhost:27017/chatbotGeminiDB4
GEMINI_API_KEY=your_key
JWT_SECRET=random_secret1234
HF_API_KEY=your_huggingface_api_key

Start Backend:

npm start

4. Terminal 2: Frontend Setup

cd client
npm install
PORT=3004 npm start

---

🏗️ Architecture

Backend (Node.js/Express)

• RAG pipeline with vector embeddings
• Gemini AI integration
• Multi-format file processing
• Podcast generation with FFmpeg + eSpeak
• Deep search engine with DuckDuckGo

Frontend (React)

• Chat UI with RAG support
• File upload manager
• Mind map visualizer
• Audio player
• Deep search interface
• Mobile-friendly design

Database (MongoDB)

• Auth & session management
• Persistent chats
• File metadata
• FAISS vector store
• Search cache

---

📦 Environment Variables

.env for /server:

PORT=5005
MONGO_URI=mongodb://localhost:27017/chatbotGeminiDB4
GEMINI_API_KEY=your_key
JWT_SECRET=random_secret1234
HF_API_KEY=your_huggingface_api_key

---

📦 Key Dependencies

Backend

• @google/generative-ai
• @langchain/community
• faiss-node
• mongoose
• express
• multer
• fluent-ffmpeg
• duck-duck-scrape
• node-cache
• say

Frontend

• react
• reactflow
• dagre
• axios
• react-markdown

---

🚀 Development

Dev Mode

cd server
npm run dev

cd client
npm start

Testing

cd server
npm test

cd client
npm test

---

👥 Team Contributions

Name	GitHub Username	Contributions
Jaya Aswanth Allu	AswanthAllu	Chain of Thought, Persistent Chat History, STT/TTS, Podcast Generator, Mind Maps, RAG, Deep Search UI/UX
Solomon Matthews	7nos	Deep Search Engine , Web Search Engine , multi model integration LLMs, rate limiting , optimize prompt for better results