Each transformation is defined per column using an expression builder. Supported operations include string functions (trim, upper, lower, replace, substring, split, concat), numeric functions (round, floor, ceil, abs, basic arithmetic), date functions (parse, format, add days, difference), and conditional logic (if-else based on value tests). You can chain multiple transformations on the same column and create entirely new columns from expressions referencing existing ones.

A live preview updates as you build transformations so you can verify results before applying them to the full dataset.