Can you describe how you optimized an Azure Data Factory pipeline to reduce execution time by 20%?

Technical