My question to for you is, when going through the data science methodology, how would you as a Director process the deployment of a model and/or visualizations? Would it be to a select controlled environment or would it be available to a broader scale? In addition, how would you then process and analyze feedback from the models and/or visualizations?