Nice work! As a software / data engineer, I totally agree. I always default to SQL and then only write Spark code when needed.