Setting up a complete machine learning stack that is able to scale with the increasing amount of data could be challenging. A recent wave of the Software as a Service (SaaS) and Infrastructure as a Service (IaaS) paradigm has spilled over to the machine learning domain as well. The trend today is to move the actual data preprocessing, modeling, and prediction to cloud environments and focus on modeling tasks only.

In this section, we'll review some of the promising services offering algorithms, predictive models already train in specific domain, and environments empowering collaborative workflows in data science teams.