This study explores the potential of using Technology Computer Aided Design (TCAD) in conjunction with machine learning to assist in troubleshooting semiconductor device failures. By utilizing properly calibrated TCAD simulation models and parameters, a large number of devices with various defects and structural characteristics can be simulated. The resulting data can then be used to train machine learning algorithms to predict the defect and structural characteristics of a device based on its electrical characteristics, such as IV and CV curves.