Fine-tuning large language models (LLMs) is a technique used to adapt pre-trained models to specific tasks or domains, improving their performance by subjecting them to additional training steps on smaller datasets. This process can significantly improve the model's knowledge, compliance with output type, and overall ability to assist organizations in accomplishing specific tasks. However, fine-tuning may not be suitable for all use cases due to potential issues such as obsolescence, cost, hallucinations, and security concerns. Retrieval augmented generation (RAG) is a promising alternative that allows LLM-powered applications to access the most up-to-date information without expensive fine-tuning steps, making it easier to evaluate and update models within the organization's infrastructure. By carefully designing prompts and using RAG in conjunction with fine-tuning when necessary, organizations can deliver the best possible experience for their end users.