Análisis estratégico de un modelo de inteligencia de negocios basado en inteligencia artificial para el fortalecimiento de la gestión de cobro en empresas de servicios públicos de acueducto y alcantarillado en la costa Caribe Colombiana

Abstract

Water supply and sewerage utility companies in the Colombian Caribbean region face significant challenges in revenue collection and accounts receivable recovery, affecting their financial sustainability and the efficiency of their commercial operations. These challenges arise from both internal and external factors. Internally, organizations rely heavily on manual processes supported by Excel spreadsheets, fragmented commercial databases across different organizational units, and a strong dependence on the Information and Communications Technology (ICT) department to retrieve historical data. Externally, outdated customer contact information, payment resistance, duplicate property records, and high levels of socioeconomic and labor informality which account for 55% of the variance in payment default reduce the effectiveness of traditional mass debt collection strategies while increasing their operational costs. This research adopted a mixed-methods approach with a qualitative emphasis, using an exploratory-descriptive scope and a cross-sectional, non-experimental research design. The study was guided by the CRISP-DM methodology, specifically the Data Understanding phase, to analyze the organizational context and assess the information available for debt collection management. Data were collected through an extensive literature review, the application of the Internal Factor Evaluation (IFE) Matrix, and semi-structured interviews with ten senior managers and subject-matter experts from public utility companies in the Colombian Caribbean region. The findings reveal that, although 80% of the organizations have established a formal sequential debt collection process, the transition between its different stages generates operational delays and information loss. In addition, the analytical maturity of these organizations remains predominantly descriptive, as 60% still depend on traditional information systems and Excel-based analyses to support decision-making. Nevertheless, the results also indicate a strong institutional willingness to embrace digital transformation. Specifically, 60% of the participants identified a hybrid strategy based on Business Intelligence (BI) and Artificial Intelligence (AI) as a strategic priority, according to the experts interviewed, predictive analytics would enable more accurate customer segmentation by distinguishing between users experiencing financial hardship, occasional late payers, and customers who postpone payments, furthermore, these technologies would facilitate automated customer contact processes in compliance with Colombian Law 2300 of 2023 and personal data protection regulations (Habeas Data), while supporting autonomous cadastral data cleansing. Ultimately, their implementation would contribute to improving revenue collection, strengthening accounts receivable recovery, and enhancing the financial sustainability of public utility companies.