i think my hope is that interviews for ML engineering would focus on domain knowledge specific to ML and not intricate questions about using typescript and react, which is the sort of stuff i simply cannot bring myself to care about enough to memorize well enough to discuss well in an interview

The premise is a basic understanding of matrices and stats. Then go through a course on machine learning (supervized, unsupervised, deep and reinforcement learning). That’s for the theory. Easy if you know your math and python.

The practical side is endless tweaking of the data and the models. And keeping yourself informed about new models and techniques (throug scientific papers)

Exploring Kaggle and participating in ML competitions is an excellent way of getting hands on experience.