Keywords: Machine Learning, APIs, Tensorflow, MuleSoft, Application Networks
In the last 10 years, workplaces have witnessed an explosion of highly specialized applications, in which a typical mid-size company may have data locked up in hundreds of applications. For data scientists, this has led to both challenges integrating data from increasingly siloed sources as well as difficulties deploying AI and ML applications in a fragmented IT landscape. In this talk, we demonstrate how data scientists solve this problem through application networks, which emerge when applications are orchestrated through APIs. We provide an end-to-end example with MuleSoft and Tensorflow, showing how data scientists can tap the data in an application network to train models and then deploy models as intelligent applications back into the network.