Using LLMs in Cyber-Physical Systems for Agriculture - ZEMELA

This paper presents the idea of developing an advisory service using the capabilities of generative artificial intelligence and in particular of Large Language Model. The service will assess the risks for farmers when preparing projects under different programs, taking into account the Bulgarian legislation related to agriculture, as well as the requirements of the relevant program. The results of a feasibility analysis are summarized in the article. Furthermore, two architectural approaches are discussed. The service will be integrated in the platform for smart agriculture named ZEMELA. A brief overview of this platform is also given in the article.