In 2022 the World Health Organization (WHO) stated that premature birth was the leading cause of death in children under five years old. New tech has emerged for Neonatal Intensive Care Units (NICUs) fueled by artificial intelligence (AI) and Internet of Things (IoT) in healthcare, hoping to reduce death among newborns. Nevertheless, approaches that incorporate instrumentation within neonatal incubators with these tools are still very scarce, so to contribute to this need, a low-cost automated system based on Raspberry Pi5 was developed to monitor and evaluate the health status of premature newborns (PNs). Specifically, a real-time PID controller was integrated to manage both humidity and temperature, employing a dynamic simulation model. Further an extreme gradient reinforcement model (XGBoost) empowers the system to analyze PNs audio, thus identifying issues like discomfort, fatigue, and abdominal pain. This algorithm predicted classification with a 73% accuracy overall. Finally the data transfer happens using the MQTT (Message Queue Telemetry Transport) protocol to a cloud server that allows interactive, real-time dashboard displays and provides customizable reports through a mobile app tailored to healthcare experts.The system provides personalized clinical support, demonstrating that comprehensive approaches like this can be useful in hospital settings, complementing timely NICU care.