Member-only story
MLOps 5: Continuous Delivery
Let’s take a look at the activities in the machine learning ecosystem. In general, we can divide these activities into build activities and run activities. Build activities focus on creating and testing the model. Run activities focus on deploying, executing and monitoring the model. There are core machine learning activities in each of them. Feature engineering, model training, testing and packaging are some core ML activities on the build side. Model deployment and inference are the core activities on the run side. Knowledge and experience in core ML activities is a prerequisite for this course. Then, surrounding these core activities is MLOps, which again, can be split into build and run.
On the build side of MLOps, we have various activities like requirements management, data and training pipelines, data governance, experiment tracking and integrations. On the run side of MLOps, we have infrastructure management, deployment, serving, monitoring, and responsible AI. For this course, we will only focus on the run side of MLOps. For each activity in MLOps run, we will discuss the purpose and context for the activity. We will discuss techniques, methods…
