What Are Some of Your Worst Theories Built in AWS, ML Solutions and AWS?

A lot of the data preparation is done in the data science world. Sometimes it's better to push that into the data engineering world, because they might be able to use full scales like Redshift and so on. Using the right tool for the right job may not ever be immediately obvious. We were writing a pipe which was doing a pre-processed step of relaving all of its data before it turned the model. On about three or four days a set of 300 million rows. And it would take days to actually enable this data. Two or three days to actually enabling this data properly, based on two separate tables that you can easily link. That was what we're just really good