This study was conducted to address the identified gap in the literature regarding the factors influencing energy consumption in soilless agriculture greenhouses. The research was carried out in a fully automated greenhouse with a closed area of 50,000 m² located in the Buharkent district of Aydın, Turkiye. During the period from 2021 to 2023, which encompasses two production cycles, data on the growth of tomato seedlings, hourly electricity consumption of the greenhouse, and hourly external environmental data were recorded. A total of 63,888 data points were collected during these two production periods. These data were analyzed using the RandomTree decision tree algorithm in the Weka program to identify the factors affecting energy consumption. Subsequently, a Machine Learning model was developed using the Multilayer Perceptron (MLP) algorithm to predict energy consumption for the next five years. The study concluded that the highest energy consumption occurs in the nutrient solution system. The most significant factors affecting this consumption were determined to be the tomato harvesting period and external temperature, in that order. In the second phase, based on the estimated consumption values obtained, a solar energy system design was created to meet the greenhouse’s hourly energy requirement of 20 kW. It was calculated that a solar energy system consisting of 960 modules, each with a capacity of 500 W and a total area of 2303 m², would be sufficient for the greenhouse in question.