Introduction

The AI Customer Support System is a web-based intelligent assistant designed to help users query documents, upload files, and receive fast answers through a Retrieval-Augmented Generation (RAG) pipeline. The system combines a modern frontend with a FastAPI backend and uses Gemini for response generation. It supports secure login, PDF upload, image upload, knowledge indexing, and interactive question answering.

Project Objective:

The main objective of this project is to build a smart support platform that can:

answer user questions from uploaded PDFs

analyze uploaded images

maintain a searchable knowledge base

provide a secure login-based interface

deploy the application on the cloud for real-time access

Working of the System

The system works in four major steps:

1. The user logs in through the authentication page.

2. The user uploads a PDF or image file.

3. The backend processes the file and stores relevant information in the knowledge base.

4. When the user asks a question, the system retrieves the most relevant content and sends it to Gemini to generate the final answer.

Technology Stack

Frontend: React, Vite, Tailwind CSS

Backend: FastAPI, Python

Database / Vector Store: ChromaDB

AI Model: Gemini API

Authentication: JWT-based login

Deployment: Vercel and Render

System Architecture

The architecture includes:

A React-based frontend for user interaction

A FastAPI backend for file handling and authentication

A vector database for storing document embeddings

Gemini for answer generation

A structured upload and indexing pipeline

Conclusion

The AI Customer Support System demonstrates how RAG can be used to build an intelligent and interactive support assistant. By combining document retrieval, secure access, and Gemini-powered generation, the project delivers a modern solution for automated support and knowledge handling.