import s3fs # csv file df = pd.read_csv('s3://{bucket_name}/{path_to_file}') # parquet file df = pd.read_parquet('s3://{bucket_name}/{path_to_file}')