Ke Sun
Senior Software Engineer at Amazon
New York, New York, United States
Summary
🤩
Rockstar🎓
Top SchoolKe Sun is a senior software engineer with nine years of experience building and scaling mission-critical systems across finance and cloud at Morgan Stanley and Amazon, currently driving Amazon Ads and global payments infrastructure. He combines hands-on engineering and evolving people leadership to design ultra-scale, low-latency platforms—such as a global payments service for billions of payment methods and one of the largest VPC networks on AWS. Ke brings a strong machine learning research background (contributions to HRNet repositories for semantic segmentation and image classification) that surfaces in rigorous model and data handling skills. Known for meticulous problem-solving and resilience, he thrives where reliability, precision, and scale intersect and often bridges research-grade ML work with production-grade distributed systems.
9 years of coding experience
5 years of employment as a software developer
Master of Science (M.Sc.), Master of Science (M.Sc.) at Concordia UniversityBachelor of Engineering (BEng), Bachelor of Engineering (BEng) at Tianjin University of Science & Technology
Chinese, English, French
Github Skills (16)
vnet10
semantic-segmentation10
net10
model-driven10
model-building10
computer-vision10
pytorch10
deep-learning10
architectures10
python10
modeling10
architecture10
model-driven-development10
datasets10
image-classification10
Programming languages (3)
Github contributions (5)
HRNet/HRNet-Image-Classification
Apr 2019 - Dec 2019
Train the HRNet model on ImageNet
Role in this project:
ML Engineer
ML EngineerContributions:25 commits, 1 PR, 24 pushes in 8 months
Contributions summary:Ke's commits primarily involve modifications to the `cls_hrnet.py` file, indicating a focus on the HRNet model implementation. The changes include adding configuration files, making adjustments to the model structure, and fixing bugs. This suggests the user is actively involved in training and adapting the HRNet model for image classification tasks.
imagenetpytorchresnethigh-resolution-netbackbone
HRNet/HRNet-Semantic-Segmentation
Apr 2019 - May 2020
The OCR approach is rephrased as Segmentation Transformer: https://arxiv.org/abs/1909.11065. This is an official implementation of semantic segmentation for HRNet. https://arxiv.org/abs/1908.07919
Role in this project:ML Engineer
ML EngineerContributions:68 commits, 1 PR, 65 pushes in 1 year 1 month
Contributions summary:Ke contributed to the core model definition and training process. They modified the `seg_hrnet.py` file, which likely defines the HRNet model architecture, addressing transition layers and stage configurations. Additionally, they fixed bugs related to missing class weights in the `cityscapes.py` dataset implementation and added a variable to the `pascal_ctx.py` dataset. These changes indicate involvement in model development and data handling for semantic segmentation tasks.
pytorcharxivhigh-resolution-netsemantic-segmentationlip
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.