Corporate credit evaluation involves measuring a firm's risk to assess the recoverability of commercial paper, bonds, and similar instruments; it is essential for smooth economic activity, and its importance grows as capitalist economies become more advanced. The credit evaluation results of credit rating agencies serve as an important basis for determining conditions such as whether to extend credit and at what interest rate when the firm issues securities or when financial institutions make lending decisions. Domestic credit rating agencies have traditionally relied on experience-based evaluation rather than scientific credit evaluation models or techniques. As the importance of credit evaluation increases in an era of financial liberalization and internationalization, there is growing interest in establishing scientific credit evaluation systems. Traditional techniques for credit evaluation have included statistical methods such as multiple discriminant analysis, regression analysis, probit, and logit models, and since the late 1980s, artificial intelligence techniques such as inductive learning methods and neural network models have begun to be applied to corporate credit evaluation and bankruptcy prediction. In this study, extensive experiments were conducted using artificial neural network models, which recent research has shown to be superior to statistical models and inductive learning methods, and based on these experiments, a neural network model demonstrating relatively high predictive accuracy was programmed to develop a credit evaluation system. NICE-AI is the first corporate evaluation system in Korea developed using neural network models, and it is provided as a menu within NICE-TIPS, a commercial information service of Korea Information Service, Inc.