Toward continuous monitoring

With that, we have set up a fully automated and robust pipeline. So far, we have successfully implemented the deployment part or module in the MLOps workflow (as we discussed in Chapter 1, Fundamentals of MLOps Workflow). It is vital to monitor the deployed ML model and service in real time to understand the system's performance, as this helps maximize its business impact. One of the reasons ML projects are failing to bring value to businesses is because of the lack of trust and transparency in their decision making. Building trust into AI systems is vital these days, especially if we wish to adapt to the changing environment, regulatory frameworks, and dynamic customer needs. Continuous monitoring will enable us to monitor the ML system's performance and build trust into AIs to maximize our business value. In the next chapter, we will learn about the monitoring module in the MLOps workflow and how it facilitates continuous monitoring.