Is There a Core Component of a ML Pipeline?

In the ML ecosystem or in the ML teams in companies, we have two main sections. We have the research section where things are mainly happening on static data. Now we're moving to the second domain within that group which is the engineering phase. There's a lot of symmetries and there are some new components coming into play the engineering part. When you're implementing your pipelines, you need to take that into account because you want to reuse as much code as possible.