Hey,
Can you use a custom component in NLU for a message preprocessing like spell correction for a message send from core, or do you need to do this with the message preprocessor? How would I do that?
I would assume both is possible since processor.py from core calls NLU parsing where each component is called?