LLM Fine-Tuning / Model Customization Planner

Decides between prompting, few-shot, RAG, and fine-tuning; if fine-tuning, designs dataset, method (LoRA), and eval.