Análise de Chatbots Habilitados por Modelos de Linguagem para Resolução de Consultas em Manuais Automotivos no formato PDF

Abstract

In the advent of Industry 4.0, marked by the integration of Artificial Intelligence, the value of autonomous systems such as chatbots, capable of responding to inquiries based on PDF files, is highlighted. In the automotive sphere, these chatbots carry the potential to reshape customer support and post-sale services, bringing optimization and conservation of time and assets. This research aimed to inspect and gauge three chatbots solutions in this scenario, focusing on questions and answers arising from vehicle manuals. Three criteria were applied: accuracy of responses, expense, and user experience. The findings revealed that the choice of approach depends on the criteria adopted, with Approach 2, ”Ask your PDF”, being the most equitable, effectively harmonizing cost, user experience, and accuracy of responses. Therefore, it is inferred that it is feasible to provide insights about the effectiveness and applicability of these chatbots in the automotive industry, facilitating the implementation of productive generative AI strategies that satisfy the sector’s demands.