i wanna code more #datascience
Ds don’t normally write production level code or implement the algorithm optimally. ML eng spend majority of their time on coding these. Good DS deals mostly with data and finding the best model/technique.
ML engineers do not make many decisions, they implement what a ds or rs built, but they are payed handsomely. Applied scientists are like ds but they implement what they build, thus make more than ds. As you can see ds make the least and are becoming a commodity. I don’t think ds is a great role to get into right now
What languages do ML engineers use? I know DS is mostly python, but for prod models do they need to optimize in something faster? Go, C++?
Depends hugely on what is being productionized Java, python, and scala +lots of cloud stuff like lamdas
Huge difference. Data scientists play with the models, perform analytics and come up with the eventually accepted model. ML engineers productionalize the eventual model, make sure the code is testable, reproducible and testable. It’s the difference between a developer and an engineer. However some people can straddle both the roles.
One of them can build products, the other one doesn't