APLICACIÓN WEB PARA LA GESTIÓN INTEGRAL Y LA PREDICCIÓN DEL IMPACTO DEL RIESGO EN LOS PROCESOS DE NEGOCIO DE LAS PYMES DE TUNJA - BOYACÁ

Abstract

The effective application of business process management enables companies to meet their objectives and goals. In addition, process management facilitates the detection of risks that may affect an organization's performance. Risks, as scenarios of uncertainty, need to be managed in order to mitigate or eliminate their negative effects. This management is a systematic process that is applied to the company's life cycle to facilitate decision-making.

In meetings held throughout this project and surveys conducted with SMEs in the city of Tunja and other municipalities and cities, various technological and organizational business risks were identified. In many cases, there is a lack of strategic objectives and no action is taken to identify and address risks. This lack of follow-up acts as a key target for the weakness in the foundations of institutions, as it is not customary to keep a historical record of the problems encountered in the processes, or the control is archaic in nature with no documentation. In other cases, documentation, risk matrices, and risk classification tables were available in Excel spreadsheets, but ultimately they were not comprehensive enough to manage risk as stipulated by institutions such as ICONTEC. With limited control over situations that entail risks, companies are exposed to a latent danger of economic and reputational losses and even the ultimate consequence of inadequate process management: business liquidation.

Comprehensive risk management is an excellent strategy for avoiding business liquidation, which is very common among SMEs, especially in medium-sized and small cities. Tunja, Boyacá is a medium-sized city with interesting urban growth indicators. However, its business and industrial fabric is not growing at the same rate; although its SMEs are formally established, they lack adequate management of their internal processes. Therefore, it is necessary to design and implement a comprehensive risk management computer application that detects, diagnoses, and forecasts the risks inherent in business processes for subsequent treatment.

This research project applies engineering methodologies, techniques, and strategies to develop this computer tool with two main components: risk detection and assessment, and risk impact prediction. During the construction of the tool, its impact was evaluated on a sample population of SMEs in the city of Tunja, using statistical instruments that allowed its reception to be analyzed. One of the scientific contributions of this research is the consolidation of a risk dataset that was constructed from the surveys conducted. This was the dataset used to train different machine learning models in order to choose the one that generated the highest accuracy. It should be noted that the data was supplemented with synthetic data generated using artificial intelligence techniques. This combination allowed the information base to be expanded and improved the accuracy of the selected predictive model.