Designing a simple AI-powered chatbot for customer support : a user centric approach
Adigun, Arafat (2024)
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-2024121034439
https://urn.fi/URN:NBN:fi:amk-2024121034439
Tiivistelmä
This thesis explores the development of an AI-Powered chatbot designed to enhance customer service on WhatsApp Business. In short, the core concept lies in the development of a chatbot that can handle customer queries by responding aptly to improve user experience without human intervention in routine interactions. This thesis work also focused on the evaluation of a chatbot for efficiency in responses and solving queries by adopting a user-centric approach.
This thesis was supported by the theoretical framework of literature on chatbot technologies and NLP in customer service automation. The review of varied case studies has shown the existing limitations and the points which have the scope for improvement within the functionality of the chatbots.
A conversational chatbot was prototyped, tested, and enhanced through user feedback surveys and performance evaluation mechanisms. Certain methodologies undertaken in the research study involved the collection and analysis of real data with respect to user satisfaction about chatbot responses and their accuracy, among other factors. Performance was measured on how well it understands natural language queries and responds with relevant information, handled multiple scenarios around customer service.
The results showed that the bot could solve some simple inquiries, reduce response time, and raise efficiency in general, while several challenges were related to the inability of the chatbot to answer more complicated queries and give personalized responses. It was concluded that while the chatbot has done an excellent job for simple-level customer service, further development of the technology could be channelled toward tackling higher-level cases and allowing greater degrees of personalization to enhance user friendliness.
This thesis was supported by the theoretical framework of literature on chatbot technologies and NLP in customer service automation. The review of varied case studies has shown the existing limitations and the points which have the scope for improvement within the functionality of the chatbots.
A conversational chatbot was prototyped, tested, and enhanced through user feedback surveys and performance evaluation mechanisms. Certain methodologies undertaken in the research study involved the collection and analysis of real data with respect to user satisfaction about chatbot responses and their accuracy, among other factors. Performance was measured on how well it understands natural language queries and responds with relevant information, handled multiple scenarios around customer service.
The results showed that the bot could solve some simple inquiries, reduce response time, and raise efficiency in general, while several challenges were related to the inability of the chatbot to answer more complicated queries and give personalized responses. It was concluded that while the chatbot has done an excellent job for simple-level customer service, further development of the technology could be channelled toward tackling higher-level cases and allowing greater degrees of personalization to enhance user friendliness.