Development of an Integrated Chatbot on the Website Using IBM Watson Assistant

This research develops an integrated chatbot using IBM Watson Assistant to improve customer interaction on the Martindo Fine Foods website. The main problem addressed is the inefficiency and inconsistency of customer service responses. Using the Waterfall method, the study followed five systematic stages: requirements analysis, design, implementation, verification, and maintenance. The chatbot, featuring a hybrid navigation system, was evaluated using comprehensive Blackbox testing across 12 scenarios. With an accuracy of 91.67%, 11 scenarios succeeded, while 1 scenario failed due to typographical errors. Over one week, the chatbot successfully handled 50 out of 51 interactions, the results show that the chatbot significantly enhances response speed and reduces unanswered messages.