The process of labeling data to train or fine-tune Large Language Models (LLMs).
A company annotating customer reviews to train a LLM for sentiment analysis