Diseño de un chatbot de preguntas frecuentes para estudiantes de postgrados en procesos de titulación de una institución de educación superior

Abstract

CONDITION FOR PUBLICATION OF PROJECT. project designs a chatbot prototype to automate answers to frequently asked questions from graduate students in the process of graduating from an Ecuadorian public higher education institution. It is based on interviews, surveys and BPMN modeling to identify 25 recurring queries and select rule-based technologies, while integrating with Telegram through Dialogflow and a webhook in Render. Objectives include: identifying FAQs through interviews with administrative staff, analyzing technologies for the chatbot architecture, and designing a prototype integrated with Telegram. The methodology employed a non-experimental approach with BPMN modeling (AS-IS/TO-BE), selection of Dialogflow using a weighted matrix, and testing with simulated data. The results show: 1) 100% precision in general responses, 2) reduction of attention times from 72 hours to less than 2 minutes, and 3) user satisfaction regarding the chatbot’s usability. It is concluded that the prototype meets the objectives of automating the management of frequent questions, freeing administrative staff from repetitive tasks to focus on performing other activities, and providing reliable responses. This optimization not only increases the efficiency of administrative work, but also improves the student experience in the graduation process by offering continuous support in this academic stage. Keywords: Academic chatbot, Automation, Degree, Dialogflow, BPMN.