In a scenario where you are working with a large dataset in Azure Databricks, you have a DataFrame 'df' with columns 'id', 'name', and 'age'. Your task is to optimize the data processing by filtering out rows where 'age' is NULL and then counting the number of remaining rows. Considering the importance of performance and correctness in a production environment, which of the following Spark SQL queries correctly creates a Common Table Expression (CTE) named 'filtered_data' that filters out rows where 'age' is NULL and then uses the CTE to count the number of rows? Choose the single best option.