AI-Generated Surf Reports: Integrating Spot Analysis and Personalized Recommendations with Generative AI
Year: 2025
Type: dissertation
Abstract: As surf forecasting becomes increasingly data-driven, the ability to generate personalized and contextually relevant surf reports using generative AI presents new opportunities for improving the surf planning experience. This thesis introduces a generative AI system that automatically produces tailored surf reports in natural language based on user preferences and live forecast data. The system combines surf spot descriptions, user intent, and environmental conditions to create informative recommendations. It integrates a vector database for semantic retrieval, real-time oceanographic data from the IPMA API, and a large language model to produce coherent and relevant outputs. Different configurations were tested, including variations in language models, embedding models, retrieval depth, and temperature settings. Feedback from experienced surfers confirmed the practical potential of the system, highlighting its ability to translate user intent into useful surf recommendations. This work demonstrates that generative AI can be effectively applied to surf forecasting and lays the groundwork for future tools that offer intelligent, personalized surf guidance.
Author: Laechelin, Maximilian
Language: English
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7111922573