The core idea of the application is to test the performance of complex behaviors. There should be a section in the application (or even a whole new application) showing a dashboard with all the metrics of the evaluation of the current model being tested. Related with #9
