Thankfully, this rewriting has never been my job. Companies where I've seen it were typically targeting Spark as the execution environment, so the production la

Thankfully, this rewriting has never been my job. Companies where I've seen it were typically targeting Spark as the execution environment, so the production languages were Java and Scala, at a ratio of about 2:1. PySpark in production was either disallowed up front or quickly disallowed retroactively after experiencing the magic and delight of data scientists shipping production PySpark.