My experience at big tech is that software engineers kinda do it all. It's almost DevOpsPmDbaDsDeMle. Where Ds is data scientist, De is data engineer and Mle is machine learning engineer.
And ML scientists are just SE with PhDs in ML.
Basically, they tend to try and not create seperate isolated roles in specialized domains, but instead hire people who can do and figure it all.
The opposite seems true in smaller companies. They break things up, try to hire different roles, specialist in different domains, and segregate each one from one another.
Obviously, mileage may vary, and I'm doing a massive generalisation of big and small businesses, but this is just my observation.
And ML scientists are just SE with PhDs in ML.
Basically, they tend to try and not create seperate isolated roles in specialized domains, but instead hire people who can do and figure it all.
The opposite seems true in smaller companies. They break things up, try to hire different roles, specialist in different domains, and segregate each one from one another.
Obviously, mileage may vary, and I'm doing a massive generalisation of big and small businesses, but this is just my observation.