At home fitness has rapidly risen recently due to the COVID-19 pandemic and stay-at-home-orders. This also produced a large set of first time users of gym equipment and structured exercise routines. Access to professional fitness trainers to assist beginners in proper exercise form has become increasingly difficult. According to the National Safety Council (NSC), approximately 468,000 injuries occurred due to exercise in 2019 before the pandemic. Without proper guidance, this statistic is bound to increase. Therefore, there is a need for systems to monitor exercise performance for both short term and long term injury prevention. We present a novel mobile app called Verum Fitness which will use the camera from a smart phone to record a user performing an exercise. Then, the app will skeletonize the user, extract angles from specific joints, and feed this data into a Fuzzy Inference System (FIS), an inherently explainable model, to classify exercise performance. With the FIS, we can provide a description of each repetition performed to determine if it could cause injury and how to improve. From our synthetically generated data, we show a training and test Accuracy of 80.42% and 71.67%, respectively, as well as high Sensitivity and Specificity for the goblet squat.