The problem I want to solve:

On-premise I have a ton of complicated SQL code in my gold layer using temp tables for intermediate results. No way around that.

I want to migrate the gold layer to DLT. I thought the best way to do this, is to use parameterized spark.sql() instead of creating a temporary view everytime for those intermediate results.

Also im outsourcing each tables code into .py files and then importing them inside the DLT notebook, to not blow up the notebook with thousands of lines of code.

Is there a better approach for comlpicated gold layer logic ind DLT?