Hi everyone! Yesterday I found this blog post which summarized the paper on the front of HN. I found this paper extremely interesting about their way to adopt ML/DL in production and the framework they are using to set up the experiments. Really curious to know your thoughts.