Fine-tuning
Fine-tuning is additional training that adjusts a model on a smaller, specific set of examples after its main training.
What it tells you about a developer
Fine-tuning experience is less common than prompt work and points to a developer who has hit a problem a prompt could not solve. What they fine-tuned and why shows how deep the AI work went.
Fine-tuning gets its name from the idea of a fine adjustment on top of something already built, the way a piano gets tuned rather than rebuilt. A model gets its broad abilities from its original training, then a developer fine-tunes it on a narrower set of examples so it performs better on one particular task or in one particular style. Fine-tuning costs more and comes up less often in everyday development than writing a good prompt, and most engineers reach for a prompt first.