EngineTalk: supported maintenance of diesel engines aboard navy ships through a RAG enabled LLM system

Over the past years, Large Language Models (LLMs) have matured to a level wherein adoption for naval applications is eligible for serious consideration. In this paper, we describe the implementation of an LLMdriven AI agent for the support of engine maintenance crew in diagnosing and finding solutions to engineering problems. By implementing a RAG component, in combination with various prompt engineering techniques, early results show that we can save significant time and resources for engine operators. Additionally, the implementation of a comprehensive evaluation framework enables avenues for reliable performance indicators.