Diseño conversacional estructurado en reglas con modelado de procesos para soporte técnico de primer nivel en una empresa eléctrica

Abstract

CONDITION FOR PUBLICATION OF PROJECT. project aims to improve the efficiency of first-level technical support in an electrical company by designing a rule-based conversational system and business process modeling (BPMN). The identified issues include a high volume of daily tickets (60-80), lack of standardization, and reliance on tacit knowledge, which negatively impact service quality. The proposed solution integrates a chatbot to automate repetitive tasks, optimizing resources and reducing response times. The project was developed in five phases: diagnosis of the current process (AS-IS), modeling and root cause analysis, design of the future process (TO-BE), conceptual design of the conversational system, and validation. Techniques such as surveys (120 users), technician interviews, and BPMN tools were used to visualize workflows. Results showed that 66.7% of technicians' time is spent on repetitive incidents, such as password resets and network issues, justifying automation. The chatbot design prioritized high-frequency use cases, with 72% user acceptance. It is concluded that implementing the conversational system will significantly enhance operational efficiency, user satisfaction, and performance measurement through KPIs like TTR and CSAT. A controlled pilot is recommended to validate results before full deployment. Keywords: Chatbot, BPMN, automation, operational efficiency.