how much math is used day to day as a data scientist/ data engineer at a big name tech company?

is it just linear regression and classification?
is your day spent cleaning data?
do you have to build ML pipelines often?