Resilience: Development of an AI based rating to increase crisis-immunity

In times of crisis, it becomes clear which companies are resilient, survive the crisis or even still use it as an opportunity for something new. Unfortunately, it also becomes clear that many companies do not succeed in this. The question arises as to what the former do differently or better. The planned research project "Entrepreneurial Resilience" will address precisely this question. The initial data basis will be created within the framework of an empirical survey. Potential levers and indicators for corporate resilience will be identified. In this context, both internal and external data will be taken into account. With the help of statistical analyses (especially regression and factor analyses), their relevance will be determined on the basis of historical figures. A neural network is formed with the most significant variables. Machine learning methods (especially supervised learning and reinforcement learning) are used to increase the predictive power of this model. The aim is to optimize the output variables, i.e. the resilience indicators, or at least to bring them above defined threshold values, so that the company is considered to be resilient. In order to validate this empirically, a benchmarking database is to be set up on the basis of the above-mentioned survey, which will contain data from both resilient and non-resilient companies in order to define the threshold values in this way.
The implementation partner valantic plans to integrate this know-how into its existing BI toolset and offer it to its customers. In the process, a light version will be available free of charge on the website. The goal, however, is to perform a comprehensive AI resilience rating on customers and advise them on how to increase their resilience. By combining empirical data with machine learning methods, this will create a unique offering on the market.