How would you handle null values, deduplicate records, and normalize inconsistent text fields in a healthcare dataset?
