🏫 University Chatbot – RAG-based Question Answering System

A powerful Retrieval-Augmented Generation (RAG) based chatbot that answers questions related to university programs, admissions, deadlines, and campus information.

Built with:

• LangChain + Ollama for LLM orchestration
• Vector DBs for semantic context retrieval
• vLLM for scalable, concurrent inference
• React + Node.js for a full-stack web interface

---

✳️ Key Features

• ❯ Context-aware question answering using RAG
• ❯ Fast semantic search with vector embeddings
• ❯ Clean and scalable architecture (React + Node.js + Python)
• ❯ Optimized for performance with caching and vLLM
• ❯ Fully offline/local inference with Ollama support

---

🧠 System Architecture

  ┌────────────┐
  │  Frontend  │ ◄──────────── User Interface (React)
  └─────┬──────┘
        │
        ▼
  ┌────────────┐
  │ Node.js API│ ◄──────────── Frontend sends user queries
  └─────┬──────┘
        │
        ▼
  ┌────────────────────────────┐
  │ Python Backend (LangChain) │ ◄──── Embeds query & retrieves context
  └─────┬──────────────────────┘
        │
        ▼
  ┌───────────────┐
  │ Vector DB     │ ◄──── Semantic search (FAISS / Chroma)
  └───────────────┘
        │
        ▼
  ┌──────────────┐
  │ LLM (vLLM /  │ ◄──── Prompt + context goes to the model
  │ Ollama)      │
  └──────────────┘
        │
        ▼
  ◄─────┐ Returns final answer
  ┌────────────┐
  │  Frontend  │
  └────────────┘