Loading

MLEngineer&Researcherbuildinginterpretableandreliableautonomy.

I work at the intersection of machine learning, continual learning, robotics, and computer vision. Focus on autonomous systems that generalize and can be understood.

Location

San Diego, CA / Hong Kong

Education

Math-CS, UC San Diego

Focus

Machine Learning · Robotics · Computer Vision

Languages

English · Cantonese · Mandarin · Japanese

Work Experience

  • ML Project Lead

    E4E Lab · UC San Diego

    Leading a 10+ member cross-functional team in collaboration with Scripps Institution of Oceanography to map 15M hectares of global mangroves via an automated Google Earth Engine pipeline and a global semantic segmentation network.

    Google Earth EngineAlphaEarthSemantic SegmentationK-MeansPython

    E4E Lab

  • Machine Learning Engineer

    E4E Lab · UC San Diego

    Architected a custom gated ensemble network fusing state-of-the-art models and built a highly reproducible PyTorch training pipeline to process 500+ GB of high-resolution drone imagery.

    PyTorchSegFormerDeepLabV3+CVATActive Learning

    E4E Lab

  • Machine Learning Research Intern

    Centre for Artificial Intelligence and Robotics (CAIR) · HKISI

    Co-authored a continual learning paper submitted to COLT 2026, proposing a dimension expansion algorithm and Newton-type optimization method with proven non-asymptotic error bounds.

    Continual LearningMathematical OptimizationAlgorithm DesignCOLT 2026

    Centre for Artificial Intelligence and Robotics (CAIR)

  • Computer Vision Engineer Intern

    SKAI · Hong Kong SAR

    Built and deployed an end-to-end YOLOv11 vertebrae detection system achieving 97.88% mAP50, processing 1,000+ diagnostic images at Queen Elizabeth Hospital via a custom GCP MLOps pipeline.

    YOLOv11Computer VisionGCPONNX.jsDocker

    SKAI