Application of the API-Based Gemini AI Model in Predicting Harvest Accuracy and Distribution of Horticultural Result

Asyhari, Ahmad
Azhari, Diah
Arif, Hilda Meisya
Nizar, Fahrul Ikhram

SINTA Journal (Science, Technology, and Agricultural)
Vol. 6 No. 1 (2025)

This study explores the integration of the Gemini AI model as an expert system to enhance the accuracy of harvest predictions and the distribution of horticultural products. The main contribution of this research lies in the application of the cutting-edge Gemini AI, which processes real-time environmental data—such as weather conditions, soil moisture, and market demand trends—to generate more accurate harvest time predictions and automated product quality evaluations. Implementation results show a 22% improvement in harvest prediction accuracy compared to conventional methods, an 18% increase in operational efficiency in distribution, and a 15% reduction in post-harvest waste. These findings suggest that AI-based expert systems offer adaptive solutions to the challenges of horticultural crop management and represent a significant innovation in modern agricultural practices.