Watch this video on YouTube
What is an effective method to preprocess prompts using custom code before sending them to an LLM?