Machine Learning Real-Time

Trần Trung Đức

Nguyen Van Huong, Thao Dien, Ho Chi Minh • ***************@*****.*** • 096******* • Github • Linkdln Education

University Of Science

Electronics and Telecommunication GPA: 3.4/4.0 2021 - 2025 Relevant Coursework: Embedded Systems, Microcontroller Programming, Digital Signal Processing, Data Structures and Algorithms, Machine Learning, Deep Learning

Experience

AI Engineer Intern – ITR VN, HCMC. Nov 2024 – Jan 2025

• Built real-time blood pressure monitoring device using STM32 and MAX30102 sensor

• Implemented PPG2BP-Net achieving clinical accuracy: SBP (MAE: 9.76 8.99 mmHg), DBP (MAE: 5.25 5.26 mmHg)

• Developed Bluetooth module and mobile app for wireless data visualization Data Annotation Intern – SNOW VIETNAM Corporation, HCMC April 2025 – Present

• Labeled high-quality datasets for ML training (500+ images/day) with >95% accuracy

• Collaborated with teams to improve AI model performance through refined data quality Technical Skills & Projects

Programming Languages: C, C++, Python, SQL, MATLAB, Verilog, Java, Kotlin Embedded Systems: STM32, GPIO, I2C, UART programming AI/ML Tools: TensorFlow, Keras, Signal Processing

Development Tools: Git, MATLAB, Mobile App Development Others: OOP, Bluetooth Communication, IoT Systems

Key Projects

Cuffless Blood Pressure Monitoring System using PPG Signal Processing

• Developed an IoT-based continuous blood pressure monitoring system using STM32 and MAX30102 sensor

• Implemented PPG2BP-Net deep learning model achieving clinical-grade accuracy: SBP (MAE: 9.76 8.99 mmHg), DBP (MAE: 5.25 5.26 mmHg), MAP (MAE: 6.26 5.89 mmHg)

• Designed mobile app for real-time visualization via Bluetooth communication

• Project selected for "Biomedical Engineering Innovation Competition 2025"

• Achieved 90% user satisfaction in testing phase with 20+ participants Real-Time Health Monitoring Mobile Application - Android

• Developed full-stack Android application (500+ lines) with Camera2 API for real-time physiological signal capture

• Implemented two-stage neural network (TF Lite) with memory optimization and background thread management

• Built comprehensive data pipeline: real-time processing AI prediction cloud sync (Google Drive API)

• Engineered 1024-point sliding window buffer management with signal normalization and preprocessing

• Integrated 85MB AI models with production-ready architecture and OAuth authentication Certifications

• Machine Learning with Python Coursera

• Deep Learning & Neural Networks with Keras Coursera

• Building Deep Learning Models with Keras and TensorFlow Coursera Awards & Recognition

• Biomedical Engineering Innovation Competition 2025

• IoT Blood Pressure Monitoring System - Team Lead

• Real-Time Health Monitoring App - Individual Developer Research Interests

Biomedical Signal Processing • IoT Healthcare Systems • Deep Learning Applications • Embedded AI Systems

Contact this candidate