Convert/compress SQL database to parquet format for RedShift

Question

I have 3 SQL databases in a s3 bucket on AWS that I would like to upload to redshift. I learned that converting them to some big data format like parquet would be much bette