Esther Kim
Manager at 삼정KPMG (KPMG Korea)
Seoul, South Korea
Summary
🤩
Rockstar🎓
Top SchoolEsther Kim is a finance and payroll manager based in Seoul with 15+ years of experience operating at the intersection of payroll operations, tax compliance, and financial reporting for multinational firms. She has overseen Korea payroll and salary arrangements for up to 1,600 employees, led annual year-end tax settlements multiple times, and managed statutory tax, audit preparation, and retirement/severance processes across APJC programs. Comfortable with systems like SAP HR, ADP, and Workday, she has driven HR-projects including M&A integrations and APJC program launches while maintaining tight control of allowances, prepayments, and closing cycles. Unusually for a payroll leader, she also contributes as a back-end/ML engineer to Intel’s high-profile OpenVINO open-source toolkit, optimizing GPU kernels and dynamic model support. That blend of deep operational finance expertise and hands-on technical contribution makes her adept at translating complex payroll requirements into reliable, auditable systems. She brings rigorous compliance focus, cross-functional stakeholder communication, and a practical knack for process modernization.
4 years of coding experience
학사, 수학 및 경영, 학사, 수학 및 경영 at 아주대학교
English
Github Skills (13)
kernel10
c-language10
cprogramming-language10
gpu10
performance-optimization10
intel10
kernel-mode10
openvino10
machine-learning9
inference9
deep-learning8
computer-vision8
python5
Programming languages (3)
Github contributions (5)
openvinotoolkit/openvino
Mar 2022 - Jan 2023
OpenVINO™ is an open-source toolkit for optimizing and deploying AI inference
Role in this project:
Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:479 reviews, 29 commits, 145 PRs in 9 months
Contributions summary:Esther's contributions primarily focused on enhancing the OpenVINO toolkit, specifically within the Intel GPU plugin. The user implemented and optimized kernels for the Intel GPU, supporting various operations such as prelu, group normalization and different activation functions. The user also worked on addressing accuracy issues by scaling activations and incorporated support for dynamic models and model caching, indicating optimization efforts for improved performance.
inference-enginepytorchmodel-optimizerdeep-learninggpu
e-ddykim/openvino.genai
Aug 2024 - Mar 2025
Run Generative AI models using native OpenVINO C++ API
Contributions:33 pushes, 1 branch in 7 months
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.