Your question is Feature Engineering for Text Classification. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
You are working on a text classification problem where raw customer messages need to be turned into features before training a model. The data is noisy, short, and full of product names, abbreviations, and spelling mistakes.
What are the best practices for feature engineering in natural language processing?