• Assist in a fast-paced startup to rapidly adapt to evolving requirements. Foster cross-functional collaboration and contribute to proposals for fundraising.• Design a wearable hardware system that is capable of monitoring Electrocardiography (ECG), Photoplethysmography (PPG), Respiration, Temperature, and Cervical Electroneurography (ENG), equipped with novel signal processing and machine learning capabilities for real-time inflammation risk prediction.• Propose integration plans and conduct quality tests on hardware and software components to ensure a reliable and robust device performance in clinical environments.

• Designed signal processing and machine learning pipelines, reducing the average detection time for potential systemic inflammation risk by 40% using non-invasive physiological and neurological measurements.• Developed algorithms for classifying human hypo/hyper inflammation biotypes from patient’s resting state data and achieved 88% accuracy in pre-clinical model.• Manufactured a bedside vital sign monitoring system and designed a graphic user interface (GUI) that actively collects physiological signals and computes real-time heart rate variability changes.• Employed optically pumped magnetometers and surface electrodes to noninvasively measure peripheral magneto-neurography and electro-neurography from 50+ subjects.• Designed spike detection and clustering algorithms to classify neural activities from autonomic nervous system, linking them with blood cytokine concentration during human Lipopolysaccharide challenge.

• Developed a noninvasive brain-computer-interface using Magnetoencephalogram (MEG) to classify Rock-Paper-Scissor gestures using deep learning methods. • Designed the MEG-RPSnet neural network architecture and achieved 85.56% classification accuracy in 12 subjects. • Operated Magnetoencephalogram (MEG) and Magnetic Resonance Imaging (MRI) for subject data collection and applied signal processing pipelines to eliminate external noises, thereby enhancing classification accuracy.

• Delivered lectures on essential machine learning topics to more than 80 students from 9 universities in South Korea.• Led project-based sessions on EDA, feature engineering, machine learning, and model evaluation techniques in KNIME and Python environments.• Facilitated the development of students’ skills in presenting research findings and writing scientific reports.