Summary
In this chapter, you learned how to deploy a TensorFlow SavedModel. This is by no means the most common method to use in enterprise deployment. In an enterprise deployment scenario, many factors determine how the deployment pipeline should be built, and depending on the use cases, it can quickly diverge in terms of deployment patterns and choices from there. For example, some organizations use AirFlow as their orchestration tool, and some may prefer KubeFlow, while many others still use Jenkins.
The goal of this book is to show you how to leverage the latest and most reliable implementation of TensorFlow Enterprise from a data scientist/machine learning model builder's perspective.
From here, depending on your interests or priorities, you may take up what you learned in this book and pursue many other topics, such as MLOps, model orchestration, drift monitoring, and redeployment. These are some of the important topics in any enterprise machine learning discussions...
