Fine-tuning
Taking an already-trained AI model and training it further on your own data.
Instead of training a new LLM from scratch (costs billions and takes months), you start with an existing model and knead it with your examples — maybe 1,000 company emails to teach it your tone, or 500 support tickets to teach it how your company answers.
For many use cases, RAG is cheaper and more flexible than fine-tuning. Fine-tuning is good when you want to change the model's style or specialise it on a narrow domain.