Donguk Jung
Software Engineer at 스켈터랩스
Seoul, South Korea
Summary
🤩
Rockstar🎓
Top SchoolDonguk Jung is a software engineer with 11 years of experience bridging low-level systems engineering and machine learning, currently focused on speech recognition and model compression at Skelter Labs in Seoul. He holds dual bachelor’s degrees in Computer Science and Mathematical Sciences from KAIST and combines strong C++ and Python skills with a deep interest in algorithms and ML/DL. At Samsung and Skelter Labs he improved on-device ASR accuracy and compressed transformer models to run on low-spec hardware, while also building robust data pipelines and optimizing decoding performance. His open-source contributions to high-profile runtimes like dotnet/runtime and the Dart SDK highlight expertise in system-level optimizations and compiler backend improvements that reduce memory and startup overhead. Fast-learning and detail-oriented, he pairs practical production experience with academic rigor to tackle performance-critical problems end-to-end. An uncommon strength is his track record of squeezing large models and runtimes into constrained environments without sacrificing accuracy.
11 years of coding experience
3 years of employment as a software developer
학사, 전산학부, 학사, 전산학부 at 한국과학기술원(KAIST)
Korean, English
Stackoverflow
41reputation
7kreached
0answers
1question
Github Skills (26)
c-language10
system-programming10
architecture10
net10
low-level10
dotnet10
asp-net10
low-level-programming10
performance-optimization10
compiler-optimization10
arm10
dart10
architectures10
cprogramming-language10
multiplatform9
Programming languages (6)
Github contributions (5)
dotnet/runtime
Jul 2019 - Mar 2022
.NET is a cross-platform runtime for cloud, mobile, desktop, and IoT apps.
Role in this project:
Back-end Developer
Back-end DeveloperContributions:423 reviews, 27 commits, 91 PRs in 2 years 8 months
Contributions summary:Donguk contributed to the .NET runtime, specifically focusing on low-level system-level optimizations. Their work includes marking relocation sections as "NotNeeded" to reduce memory footprint, enabling NGEN for aggressively optimized methods to improve startup time, and converting AssemblyNative::Load and AppDomainNative::CreateDynamicAssembly to QCalls to remove libunwind overhead. They also fixed build errors and addressed various platform-specific issues.
dotnetruntimelinuxcsharpxamarin
dart-lang/sdk
Apr 2021 - Jun 2021
The Dart SDK, including the VM, JS and Wasm compilers, analysis, core libraries, and more.
Role in this project:Back-end Developer
Back-end DeveloperContributions:5 commits, 3 PRs, 7 comments in 1 month
Contributions summary:Donguk primarily contributed to the Dart SDK's compiler and runtime, focusing on optimization and code generation. Their work involved modifying the ARM backend, specifically updating instructions to use immediate operands for arithmetic operations. Additionally, the user addressed issues related to printing and improving the accuracy of compiler output, such as stack bitmap printing. The commits also included refactoring and code improvements to enhance performance.
dartsdkdart2jscore-librariesflutter
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.