In a world that becomes ever more reliant on the latest technologies, intelligent automation is rapidly reshaping various sectors, with the field of renewable energy leading the automation revolution. This research proposes the design and implementation of an intelligent online solar tracker system based on artificial intelligence (AI), which automatically controls the orientation of solar cells in an embedded system. The proposed work uses machine learning (ML), which is a subset of AI. Three models of machine learning (ML), including Linear Regression, Decision Tree, and XGBoost models, were designed, implemented, and analyzed. The experimental outcome showed that the XGBoost model performed best, recording an accuracy of an incredible 99% with a mean absolute error of 0.646 and a root mean square error of 1.08. The hardware design combines an ESP32 microcontroller board with an Arduino Uno board that allows sensor actuations. Data synchronization among all the modules is done by using the real-time database in the Firebase application. The system also uses a corresponding smartphone application developed by using the Flutter application framework. The combination of intelligent embedded systems with machine learning models enables this research work to showcase a cost-effective, portable, and fully automated solution that can increase the efficacy of solar energy installations. The proposed framework attempts to reveal the ever-increasing use of automation by AI models in improving sustainable energy resources.