Abstract

This paper addresses the challenge of efficiently handling a diverse array of customer queries by proposing the development of an innovative web-based customer support chatbot. The objectives encompass creating a versatile system capable of interpreting and resolving a spectrum of customer complaints, enhancing support staff efficiency, and facilitating knowledge base updates. The proposed methodology employs the MERN stack for web app development and integrates Generative AI and pre-trained Large Language Models (LLMs), specifically OpenAI's prebuilt models, for intelligent responses. The pseudo code outlines the interactions between the chatbot, support staff, and administrators, ensuring a seamless process. Results indicate the chatbot's commendable ability to accurately interpret queries, showcasing language versatility in various Indian and foreign languages. Efficient handling of varied phrasings and also accommodating grammatical errors. The use of prebuilt models is highlighted as a practical and efficient solution, reducing implementation time compared to custom models. Additionally, the implementation is recognized for its scalability, positioning it as a practical and efficient solution for evolving customer support needs in the near future.