Thanks for your response! I tried the explode function (which is a pretty dangerous name for a function) and made sure to import pyspark.sql.functions and even made a new cluster with the most updated databricks runtime version, but I still get a "explode not defined" error. I even tried importing directly pyspark.sql.functions.explode but that model couldn't be found.

I then looked into the "Querying semi-structured data in SQL" documentation. I tried some of the methods there like raw:contacts.emails[*].emailId but got errors revolving around "raw column not existing" and when removing the "raw:" part I got an "invalid usage of '*'" error.