Describe a time you had to clean a massive, unstructured dataset for Banco Santander. What tools did you use and why?