This chapter showed us how to use Engine Templates (http://predictionio.incubator.apache.org/gallery/template-gallery/) which PredictionIO provides to build a server, and an engine to perform common machine learning tasks, such as product Recommendation, lead scoring, Frequent Pattern Mining, classification, sentimental analysis, and various others. As we discovered, the Heroku platform, which is a part of the Salesforce App Cloud, can be used to host the machine learning applications built using PredictionIO. The Lightning Components on the Salesforce platform provide a powerful user interface experience to integrate the PredictionIO applications within the Salesforce application natively. Note that the code used in this chapter for the Scala part is courtesy of the PredictionIO GitHub repository and the Salesforce evangelist team. The Apex code and Lightning Component code demonstrates how one can integrate PredictionIO offered endpoints within the Salesforce lightning experience.

In the next chapter, we will explore the Einstein vision services provided by the platform for image recognition and prediction.