Is a strong understanding of mathematics necessary to become a machine learning engineer, or can someone with a strong interest in machine learning learn it over time?

A decent understand of certain kinds of mathematics is very important to become a machine learning engineer or data scientist, yet. And someone with a strong interest can acquire that background over time, certainly. Just be aware that it will take work on the side for you to get that background if you do not have it, and be aware that you may have trouble in interviews when you’re asked questions that are designed to determine if you have this background.

Machine learning is not just programming or putting data into “black box” models that you are not required to understand. You have to be able to reason about data — when a model does not work well, why does it not work well? Even something simple like a penalized linear regression — if it’s not working, what about the data is causing it to not work? Analyzing this means you need to understand certain statistical concepts, which involve some mathematics — are the data non-stationary? Is there heteroskedasticity? Is the relationship highly nonlinear and do you need a non-linear model?

If you understand concepts like those, great. If not, things like this are rock bottom basics that it is very reasonable to expect you to know, if you expect a six-figure salary in return.