Summary
In this chapter, you learned the importance of monitoring ML models deployed in production and the different aspects of models to monitor. You dove deep into multiple end-to-end architectures to build continuous monitoring, automate responses to detected data, and model issues using SageMaker Model Monitor and SageMaker Clarify. You learned how to use the various metrics and reports generated to gain insight into your data and model.
Finally, we concluded with a discussion on the best practices for configuring model monitoring. Using the concepts discussed in this chapter, you can build a comprehensive monitoring solution to meet your performance and regulatory requirements, without having to use various different third-party tools for monitoring various aspects of your model.
In the next chapter, we will introduce end-to-end ML workflows that stitch all the individual steps involved in the ML process together.
