We present our case of success in classifying user tickets by applying different techniques on natural language processing. With our model, we can determine the department of the software company to which to direct a new ticket entered by a user with an accuracy greater than 90%.