What do you like most about data, and how do you ensure data integrity when working with large, messy datasets?