To Developers, I am writing Parquet compressed data to memory and then streaming it to the cloud. I want to optimize my memory usage and be able to compress the data to the