Water quality is crucial for health and productivity in livestock operations, as contaminated water can increase disease transmission risks and lead to livestock losses, directly impacting agricultural performance and veterinary costs. Traditional water quality monitoring methods, such as laboratory tests, are costly, slow, and often unsuitable for real-time detection, limiting prompt corrective actions, particularly in agricultural settings. Emerging technologies, including AI, IoT, and Edge Computing, now enable rapid and precise contaminant detection through multi-parametric sensors and machine learning algorithms, significantly reducing detection latency and enhancing the reliability of health monitoring systems. This paper proposes an intelligent model based on artificial intelligence and Edge-AI computing for real-time detection and prediction of contamination at livestock water points. By integrating multi-parameter biosensors, our layered architecture (Perception, Edge-AI, Network, Cloud) enables optimal resource management with reduced latency and increased accuracy compared to traditional methods. Experiments demonstrate fast detection of pathogenic contaminants, thus enhancing disease prevention in agricultural environments. This innovative model, which combines IoT, artificial intelligence, and autonomous floating platforms, also has potential for environmental applications.