I know this has been asked before, but the last question is about a year or two old, and the industry moves fast, so I think a new post might be warranted.

So I've been an SWE for about 9 years now, and I've wanted to try to switch careers to become an ML Engineer. So far, I've:

* learned basic theory behind general ML and some Neural Networks

* created a very basic Neural Network with only NumPy to apply my theory knowledge

* created a basic production-oriented ML pipeline that is meant as a showcase of MLOps ability (model retrain, promotion, and deployment. just as an FYI, the model itself sucks ass )

Now I'm wondering, what else should I add to my portfolio, or skill-set/experience, before I can seriously start applying for ML Engineering positions? I've been told that the key is depth plus breadth, to show that I can engineer production grade systems while also solving applied ML problems. But I want to know what else I should do, or maybe more specifics/details. Thank you!