The increase in the demand for mass consumption products derived from the production of the livestock sector has significantly increased the impact generated by this sector in the Colombian economy. According to the Colombian Federation of Livestock Farmers (Fedegan), per capita milk consumption in Colombia is 148 liters per year and the raw milk market is 5.7 billion pesos. Likewise, it is estimated that there are 1.41 million head of cattle dedicated to milk production, a figure that shows a growth trend. This behavior shows the need to use technological tools that support the monitoring of animals and allow control of nutrition, diseases, geographical distribution and life cycle of cattle. Taking this premise into account, this work aims to develop a predictive analytical model embedded in a monitoring system based on Internet of Things (IoT) technology, focused on the raw milk production operation of Hacienda La Diana, located in the Bogotá Savannah. Specifically, a predictive analytical model is built from a monitoring system designed with georeferencing sensors, which allows predicting the behavior of cattle, in terms of: (a) diseases, (b) milk production and (c) classification. So that it supports the administration in the early detection of diseases, the optimization of herd observation times, the increase in the control of raw milk production and facilitates the geographical distribution of cattle.