Credit risk assessment plays a very decisive role, especially in banking and finance. Usually, rating agencies or credit institutions deal with credit risk analyses. As a result, the level of knowledge in this area is already very pronounced. However, the mapping of self-calculated credit risk in commodity trading is still in its beginnings. This thesis's primary aim is to develop a unique selling point in the form of a credit risk module for the Commodity Trading & Risk Management Software ComCore of the partner company ComFin Software GmbH. To achieve this, a selection of already existing computational models is examined, compared with each other, and weighed regarding their practicability to adapt it to personal interests eventually. Furthermore, the requirements for the system are worked out in the form of a requirements analysis and documented in detail. Subsequently, a basic requirement specification is prepared for the company's in-house developers. This contains all technical and non-technical requirements and a roadmap for their realization. The software module has not yet been fully validated; however, it has already completed test runs using economic figures from a sample company. The model needs to be validated in cooperation with customers to adapt or, if necessary, fine-tune the calculation process. The key findings of this thesis are the combination of three different computational methods to determine the probability of default of a counterparty independently. It has to be noted that the calculation of credit risk is a very intensive endeavor using numerous variables. This shows that critical computational concerns must be addressed before all tool parts can consistently produce reliable results. Nevertheless, this work appears to be a foundation for the future of credit risk analysis in commodity trading & risk management.