Paper Title
Automated Pdf Q and A Chatbot: Harnessing Ai for Efficient Information Retrieval

Abstract
In today’s digital age, extracting and storing in- formation from data plays an important role in increasing efficiency and productivity. This article presents an AI-driven chatbot designed to extract and answer questions from PDF files containing question-answer pairs. The system uses natural language processing techniques such as text extraction, text chunking, sentence embedding, and cosine matching to match user queries with relevant content. Additionally, generative AI models enhance the chatbot’s conversational capabilities by supporting responses with contextual understanding. This ap- plication leverages Streamlit and integrates multiple libraries such as PyPDF2, Text Converter, and Google’s GenerativeAI API to enable interactive user interaction. Evaluation of the system has shown that it provides good results in providing accurate information and integrated responses, managing information, supporting people’s consumption, and making it suitable for use in environmental education. This project, conducted in collaboration with Nokia Bangalore University (NBU), involved designing the base architecture of the chatbot and incorporating useful security features, including data ingestion prevention and log maintenance, to ensure robust and secure interactions.