Imagine being able to predict the height of ocean waves with incredible accuracy, using the power of AI. This isn't science fiction, but the reality of a groundbreaking new framework called Orca. Traditionally, predicting significant wave height (SWH) relied on complex numerical models that were computationally expensive and slow. While machine learning offered a faster alternative, it struggled with the scarcity of real-world ocean data. Orca tackles these challenges head-on by combining the strengths of large language models (LLMs) with a clever spatio-temporal awareness module. Orca works by dividing buoy data into time segments, encoding buoy locations spatially, and using carefully designed prompts to guide the LLM. This allows it to learn complex wave patterns from limited data, outperforming traditional methods and existing machine learning models. Tested in the Gulf of Mexico, Orca demonstrated remarkable accuracy in predicting SWH. This has significant implications for various maritime applications. More accurate wave predictions can improve safety at sea, optimize shipping routes, and enhance the efficiency of offshore energy operations. The speed of Orca also makes it suitable for real-time forecasting, potentially providing early warnings for dangerous wave conditions. While Orca represents a significant leap forward, the research team is already looking ahead. Future work will focus on enhancing Orca's efficiency and expanding its forecasting capabilities to predict wave heights over longer time horizons. This innovative approach signifies a new era in our ability to understand and predict the complexities of the ocean, with the potential to transform industries and improve safety for those who work and live by the sea.