KellenĀ Sunderland
Senior Software Development Engineer at Amazon Web Services (AWS)
Seattle, Washington, United States
Summary
š¤©
Rockstarš
Top SchoolKellen Sunderland is a Senior Software Development Engineer with 13 years of experience building and optimizing production ML systems at AWS, including contributions to Amazon Translate and Amazon Go. He blends deep hands-on expertise in machine learning, performance tuning, and production debugging with a strong MLOps and CI/CD background, having improved build, test, and type-checking infrastructure on high-profile open-source projects like Sockeye and Apache MXNet. His work spans from low-level performance and profiling (NVTX, OpenBLAS) to service-scale deployments, making him comfortable across research, infrastructure, and product boundaries. Earlier in his career he shipped multiple mobile apps and enterprise-grade CDN and education platforms, demonstrating a pragmatic focus on reliability and measurable outcomes. Based in Seattle, he pairs rigorous engineering discipline with a knack for improving developer workflows and documentation. Colleagues would note his curiosity about interesting technologies and his tendency to fix flaky CI and subtle production edge cases others miss.
13 years of coding experience
10 years of employment as a software developer
BSc, Computing Science, BSc, Computing Science at University of Alberta
Stackoverflow
611reputation
60kreached
9answers
12questions
Github Skills (32)
markdown10
pytorch10
machine-translation10
c-language10
python10
mxnet10
build-system10
cmake10
cicd10
mlops10
deep-learning10
markdown-it10
cprogramming-language10
documentation10
testing9
Programming languages (6)
Github contributions (5)
apache/mxnet
Apr 2017 - Nov 2019
Lightweight, Portable, Flexible Distributed/Mobile Deep Learning with Dynamic, Mutation-aware Dataflow Dep Scheduler; for Python, R, Julia, Scala, Go, Javascript and more
Role in this project:
Back-end Developer & DevOps Engineer
Back-end Developer & DevOps EngineerContributions:3 reviews, 112 commits, 249 PRs in 2 years 7 months
Contributions summary:Kellen contributed to the Apache MXNet project by addressing build and testing issues related to dependencies like OpenBLAS and ARM-based systems. They enabled specific features within CI/CD pipelines, such as clang builds and added NVTX profiling. Additionally, they worked on improving the stability of the CI environment by disabling flaky tests and managing test dependencies.
pythonschedulerdataflowmutationdata-science
awslabs/sockeye
Jun 2017 - Apr 2018
Sequence-to-sequence framework with a focus on Neural Machine Translation based on PyTorch
Role in this project:MLOps Engineer
MLOps EngineerContributions:11 commits, 18 PRs, 1 push in 10 months
Contributions summary:Kellen focused on improving the build and testing infrastructure and integrating type checking into the project. Their contributions included enabling Travis CI, integrating mypy for static type checking, and refactoring arguments. They also addressed issues such as empty input handling and checkpoint averaging to improve the model's robustness. Furthermore, they cleaned up lint issues in tests and updated the documentation.
sequencetranslationsockeyeneural-machine-translationseq2seq
Find and Hire Top DevelopersWeāve 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.