How do I deconstruct a composite function into its constituent functions using inverse functions?

Um... I don't quite understand what you mean. Could you post and example please?