This work proposes an IoT-based solution that uses ECG signals for early detection of sleep apnea, a lightweight, ECG-based apnea detection system that integrates AI to improve predictive accuracy and patient outcomes. We propose an innovative hardware-software solution designed to leverage ECG sensors and electronic components for real-time cardiac signal acquisition. Our system collects data from multiple ECG sensors, processes the signals, and transmits them to a mobile application acting as both a gateway and a display interface. Furthermore, a web application allows users to visualize their cardiac activity. By integrating these technologies, our approach enhances predictive accuracy and improves patient outcomes through a lightweight, real-time apnea detection system. Among all tested models, the CNN-based model achieved the highest accuracy, making it the preferred model for real-time prediction. A web application was developed to visualize ECG signals, provide apnea predictions, and alert users when anomalies are detected. The proposed solution demonstrates the potential of combining AI and IoT to support remote and continuous health monitoring.