We have solution implemented for ingesting binary file ( .ZIP ) into delta lake, Currently we are using the below solution within our pipeline.

1. Unzip the file and extract the XML file.
2. Parse the XML using python libraries.
3. Flatten the nested xml columns.
4. Store it to delta table.

This solution is working fine for small set of files ( 25 ). When we are processing large set of files ( 650 ) it is taking more time than expected.

Would like to know if we have a better solution to speed up the process.

Few things to note about the Xml file, This is a nested XML file which is having around 600 columns.