import awswrangler as wr wr.pandas.to_parquet( dataframe=df, path="s3://my-bucket/key/my-file.parquet" )