AI Platform of Enterprise Financial Stability Analytics

This paper introduces an AI platform designed for analyzing the financial stability of enterprises, with a specific focus on cultivating optimal decision-making capabilities. The platform enables users to construct diagnostic models utilizing logistical regression, deep learning neural networks (NN), and predicate rules. Moreover, users can develop predictive models for diverse processes using regression systems and neural network models and apply these models in real-world scenarios. This paper discusses a simulation-based mathematical model for evaluating the financial stability of small businesses. The model is implemented using FinSim, a software package developed by the authors and previously described in Munjishvili et al., (2023). The goal of the platform's general model is to achieve the following objectives: (1) develop a specific economic-mathematical model for assessing the financial stability of the enterprise and predicting bankruptcy, which can be applied in the real activity of the given enterprise; (2) learn and apply the new technology for developing an economic-mathematical model of financial stability assessment and bankruptcy forecasting for each enterprise in the study, and (3) develop organizational and technical measures for ensuring financially sustainable operations of an enterprise based on modeling the arguments included in the model of the real enterprise. In this article we apply the models to the real small-size enterprises’ financial statements data, which in line of the Georgian law of accounting belong to IV category. Using the available statistical data of three years, we carry out financial stability diagnostics of six real enterprises from IV category, namely: Nika 95 LLC, Sefo LLC, Offices LLC, Commerce LLC, Friendship LLC and Pharmakon LLC. The study finds that even with limited three-years data, the simulations predict bankruptcy.