Summary
Jay Lee is an AI research engineer and team lead with nine years of experience building high-throughput, mission-critical ML systems and production inference engines. He bridges research and engineering, having optimized Lunit’s CXR and MMG inference pipelines (up to 2.4x runtime and 31x postprocessing improvements) and led deployment-focused R&D that preserved model accuracy while drastically improving speed and reliability. Formerly a senior ML engineer at Deeping Source, he cut costs and boosted EdgeTPU performance with novel detection architectures and automation that eliminated onsite deployments. Jay’s open-source contributions (yolov5, torch2trt, DeepStream-Yolo) and personal tools (pt2keras, decko) reflect a practical focus on exportability and inference robustness, including automated verification for PyTorch→ONNX fidelity. Trained at Seoul National University’s data mining lab, he blends deep learning research—especially anomaly detection and data mining—with pragmatic system design and a knack for simplifying complex production workflows.
9 years of coding experience
9 years of employment as a software developer
Master's degree, Computer Science, Master's degree, Computer Science at Seoul National University
UNSW Sydney