Scaling Machine Learning

For analytics, modern data platform frints separate sorg and compute. So you must scale your sequal queries as much as you want. But also, ray is a general purpose substrait for scaling anything in the machine learning pipe line, including data processing. An amazon is using it internallyn amazon dot com, not a a w s. And e dat taskas also readyty like the bridge from like the pandas data stut to make them distributed.