Irrigation is an essential practice concerning the proper use of agro-resources; however, it continues to be one of our greatest challenges because citizens have been increasingly faced with water scarcity, and there is great pressure to improve the sustainable use of resources. Many water scarcity issues, caused by traditional irrigation methodologies that waste huge quantities of water and inhibit agricultural yield potential, stem from the lack of suitable prediction & real-time observation techniques. This paper aims to overcome these limitations through a low-power, affordable, and high-accuracy irrigation system based on ML- and IoT-supported technology. The methodology consists of a two-part exploratory research process. In the first stage, a variety of ML algorithms were systematically tested for accuracy in deciding irrigation action (e.g., SVM, RF, and KNN). Utilizing Kaggle for publicly available data, each model went through a thorough training and validation testing process with an emphasis on hyperparameter tuning, k-fold cross-validation, and full-fledged preprocessing. The results of the evaluation regarding accuracy (99.98%), precision (0.99), recall (0.99), F1 score (0.99), and AUC-ROC show that the RF model is the best-suited solution with high performance and stability. The next stage was the design of a novel IoT-based architecture system combining the selected ML model to make real-time decisions for irrigation. The system collected and processed field data using NodeMCU ESP8266 microcontrollers with soil moisture and DHT11 sensors continuously. The processed sensor data was sent to ThingSpeak, an IoT cloud, where the RF model gave real-time irrigation suggestions based on live sensor values. To control the irrigation pump operation using relays, these decisions were sent to a microcontroller (ESP32). Experimental validation shows that this integrated ML and IoT approach saves around 69% water and improves plant growth, as well as reducing the total consumption of freshwater more than conventional irrigation systems.