The project utilized Logistic Regression and Artificial Neural Networks to predict loan defaults for a Peer to Peer lending platform. Data from the Lending Club was used to train the models on a 5 fold cross-validation. Based on those predictions, it was tested to see if a profitable portfolio could have been constructed with these model predictions. The overall model performance would not have generated a positive profit for investors and would require additional training and fine-tuning for a production use case.