A Retrieval-Augmented based Agentic Chatbot

This project is completely based on the web and promotes the Retrieval-AugmentedGeneration (RAG) model to help users upload, manage, and search educational resources such as PDFs and previous years’ question papers. By incorporating ArtificialIntelligence (AI) and Natural Language Processing (NLP) technologies, the systemcan realize user questions and arrange clear, accurate, and context-based answers tothe questions. This makes the learning process more interactive and effective, with achatbot interface enabling users to have natural conversations while retrieving information from uploaded documents.React and Vite are used to implement the front-endof the system. These two offer a smooth and user-friendly interface for users. Theseare used to sign in securely, navigate through various sections, and view files. Thebackend is developed in Node.js and Express. Use Node.js and Express to managedata storage, file uploads, and application logic in a well-organized and scalable way.Cloudinary is being integrated into this project because of safe file storage, while adatabase is being used to manage the documents and the corresponding metadata inan efficient way. By leveraging the power of the PDF content in conjunction witha Large Language Model (LLM) using RAG, the system can create meaningful andcontext-aware responses to user queries. It also has essential features, like robusterror handling, a way of protecting access to certain pages and an intelligent searchfunction. Mainly, the given project provides a highly efficient and modernized systemof educational resource management. This project can help users find queries fromPDF. Users can upload their PDF and ask questions from it. This project deliversthe answer to the users, which available in the PDF.