The session will use a sales database to predict the likelihood of a customer to buy a bike (or another product). The modeling in Azure Machine Learning (AML) provides such an application for this prediction. The cloud-based algorithms involve a multitude of steps along the way. The analyst can use AML to decide the architecture and/or features needed for such an analysis. This includes training the model with data and tuning the model. Azure let's the analyst scale down for testing, then scale up for large data sets. Model versions can be tracked as well as using different parameters. Even though the flow looks complex, the steps are intuitive all the way to deployment to others to use.