A major advance in computerised farming technology, precision agriculture makes it possible to apply the right quantity of water and fertiliser to agricultural products at the right time to increase production. A system that effectively manages and optimises the watering of plants is referred to as an intelligent irrigation mechanism. In order to monitor and modify the irrigation process depending on the present circumstances and plant requirements, such mechanisms often contain a variety of control mechanisms, sensors, and actuators that control them. After numerous algorithms with multiple layers of regression and the addition of clustering are computed, it is shown that XGBoost combined with the k-means outperforms when compared with other algorithmic configurations that are utilised in models. This strategy is used as a conduit between the end Internet of things (IoT) gadgets and the cloud in order to reduce the amount of computationally intensive processing carried out on cloud servers. Thus, the servers linked to a local edge network execute the prescribed algorithmic calculations. An intelligent irrigation system may become more proactive, accurate, and adaptable by utilising machine learning algorithms, improving water efficiency, agricultural yields, and overall environmentally friendly resource management. On the other hand, the kXGB approach (XGBoost + k-means) yields higher accuracy and lower mean square error (MSE).