Obstructive Sleep Apnea (OSA) is a common sleep disorder that causes repeated instances of partial or complete airway closure, leading to pauses in breathing (apnea) or significant decreases in airflow (hypopneas). OSA disrupts sleep and reduces oxygen levels, which can lead to other health issues and significantly impact the quality of life across age groups. Traditional OSA detection methods, such as polysomnography, symptom assessment, questionnaires, and physical examination, are time-consuming and require experts such as doctors for diagnosis. This study presents a new approach for designing and using a portable device that combines physical parameters, such as oxygen saturation, heart rate, and chest movement, with snoring sound pattern recognition using a logistic regression machine learning algorithm. To monitor and classify sleep apnea events. The prototype device achieved an accuracy of 98% for predicting OSA. This hybrid approach offers a more convenient and effective method for detecting and monitoring OSA, potentially improving the screening and management of this common sleep disorder, and enhancing the quality of life of affected individuals.