Joined LinkedIn recently and unfortunately compared to my past experience where I did a lot of modeling using ML, the data science teams at LinkedIn are mostly focused on doing product analytics. Maybe 20% do some other things like building libraries and building pipelines for data but majority of the DSers are actually just doing product analytics which a lot of times doesn't require any ML. Which companies actually have DS roles where you can work on not just generating insights but actually building predictive models?
This is widely known isn’t it? Did you expect something else?
Sounds like a typical DS job
Dolby
TC and yoe?
Did the interview process have a round for ml?
I don’t think many people think DS do ML. At least not for the last few years. If that’s what you’re looking for then those kinds of roles are usually for ML engineers or serious academic researchers. Don’t see much DS doing that and to be fair most DS aren’t qualified for serious ML kind of work either
At least it pays just as much as ml eng right?
That seems about right and reasonable. DS’ should be paid even less than data engineers in my opinion
It shouldn’t be new to any of us now. Most days scientists are glorified analysts. People who think they’re tech because they use pandas instead of SQL lol
Data scientists in most of tech companies now don’t do production ML. That’s the job of ML engineers.
Microsoft's data scientist and Amazon's applied scientist are still doing production ML.
Which makes sense. Data science already was way too broad (analytics, causal inference, machine learning, product sense like a PM) and the smart companies figured out it didn’t make sense to try to find a unicorn. Rather, hire specialists and let people focus on one area. Startups and smaller places probably still expect DS to code like engineers and talk like PM’s