Watch this video on YouTube
A Generative AI Engineer is creating an LLM application that allows users to generate personalized birthday poems based on their names. What technique would be most effective in protecting the application against potentially harmful user inputs?