Summary
In this chapter, we introduced the concepts of data drift and target drift, and examined different approaches to performance monitoring in ML systems.
We started by introducing important concepts in the realm of performance and monitoring, different types of drift and business metrics to monitor, and the use of AWS CloudWatch as a tool to implement monitoring and alerting in real-time systems.
Performance and monitoring is an important component of our architecture, and it will allow us to conclude an important layer of our ML system's architecture. Now let's delve into the next chapter on advanced topics in MLflow.