Very interesting, thanks for the explanation. I work on an ML team as well and definitely notice the heavy reliance on analysis. Almost like reading tea leaves at some point.