IITI_BOT

Live URL for our chatbot: https://iit-indore-bot.vercel.app
WARNING ⚠️: The chatbot may take 8-10 minutes to generate responses if the backend server or docstore server (deployed on Render) is in a cold start state. It may also fail to respond due to "Pathway effect" which states that pathway can crash anytime without any changes in the previously working code. It is advisable to run the project locally.

📌 Introduction

IITI_BOT is a RAG-based AI chatbot built specifically for the IIT Indore community. It is designed to answer queries relevant to IIT Indore using official sources and context-aware response generation.

🎯 Purpose

We intend to create an intelligent, AI-powered chatbot tailored to autonomously help students, faculty, and visitors. This chatbot is designed to provide instant, reliable, and context-aware responses to a wide range of queries related to the institution.

✨ Features

The core of our chatbot is a multi-agent Retrieval-Augmented Generation (RAG) system. This means several AI agents collaborate as a team and generate a clear, helpful answer based on the most relevant documents. Our aim is to provide fast and trustworthy responses based on official IIT Indore information, improving campus experience for everyone.

📁 Repository Structure

• IITISoC_Frontend/ – Vercel-deployed frontend UI.
  Instructions to run locally
• IITISoC_Backend/ – Pathway based backend server with main backend logic and LLM orchestration.
  Instructions to run locally
• IITISoC_DocStore/ – Pathway's document store server.
  Instructions to run locally

🎥 Demo Video

Watch the demo: Google Drive - Demo Video

📄 Final Report

Read the full report: Google Drive - Project Report