Kevin Chon

Engineering Clients: Huupe, Collision Sciences, Musi (Top 50 Global iOS App), other independent entities • Musi: Designed and implemented the first production event handling data analytics pipeline and machine learning infrastructure supported on AWS, dealing with a minimum of more than 2.7 million monthly active users. Defined custom event handlers utilizing Redis for queue management, connections, spanning workers in a Python Flask environment in conjunction with AWS Kinesis & AWS Redshift for cloud data management • Real Shot (AR basketball app): Led team as Head of Machine Learning to develop a real-time object detection and classification model from custom collected basketball shooting videos data for Nreal AR glasses and Unity built game app • Collision Sciences: Implementing ML prediction system based on car sensor data, iOS app feature development • Learning is Fun (NLP App): Researched and developed a system taking in custom sets of educational articles and generating automated MC questions and answers for an iOS educational prototype app

Worked as a machine learning engineer in the Algorithm team in OPPO US R&D Seattle office. • Conducted computer vision research in 3D depth estimation and object detection • Contributed to AR translate app, particularly focusing on Korean translation use case. • Conducted time series research in real time audio signals processing for audio super resolution • Filed two patents, one related to intelligent content rendering using semantic analysis models and another related on generative pseudo lidar object detection on mobile Patent - Content rendering using semantic analysis models : https://patents.google.com/patent/WO2022250690A1 Patent - Methods and systems for generating point clouds: https://patents.google.com/patent/WO2022099225A1