Resume

Kyushick Lee

Software engineer & computer architect — LLM inference/training, AI accelerators, and systems performance.

10+ years across memory and computing system architecture, parallel programming, and system resilience, with a recent focus on LLM inference and training. Currently building the LLM serving stack for Microsoft's Maia AI accelerators; PhD in computer architecture from UT Austin.

Experience

Senior Software Engineer — Azure Hardware Architecture, AI Frameworks — Microsoft Aug 2021 – Present

Redmond, WA

Building kernels, runtime libraries, and an integrated LLM serving stack on Maia ASIC accelerators. Designed the Maia host/device programming model, delivered the SDK and PyTorch/ONNX Runtime integration, owned MoE kernels, and partnered with OpenAI for the Maia-powered GitHub Copilot demo at Ignite 2023.

Software Engineer II — Azure Hardware Architecture — Microsoft Aug 2019 – Aug 2021

Redmond, WA

Built kernel and collective libraries for an FPGA training accelerator integrated with ONNX Runtime, a hardware abstraction layer, a simulator, and checkpointing/CI infrastructure.

Graduate Research Assistant — LPH group (Advisor: Mattan Erez) — UT Austin Aug 2013 – Aug 2019

Austin, TX

Lead developer of the Containment Domains resilience runtime, analytical model, and tools across MPI, CUDA, and Legion.

Graduate Engineering Intern — Intel — Open Source Technology Center May 2018 – Aug 2018

Hillsboro, OR

Characterized Node.js front-end bottlenecks with perf and optimized code layout from profiles.

Research Intern — NVIDIA Research May 2017 – Aug 2017

Austin, TX

Studied resilience trends in GPU-dense systems and scalable GPU checkpointing.

Research Intern — NVIDIA Research May 2016 – Aug 2016

Santa Clara, CA

Built a transparent checkpointing system for CUDA programs, evaluated on HPC applications.

Research Intern — Lawrence Livermore National Laboratory Jun 2014 – Aug 2014

Livermore, CA

Built an analysis tool predicting performance and soft-error effects in applications.

Education

PhD in Electrical & Computer Engineering, The University of Texas at Austin (2019)
BS in Electrical & Computer Engineering, Hanyang University (2013)

Skills

Languages: C / C++CUDAPythonMPIOpenMPVerilog

Systems & Perf: Performance analysisVectorizationGPU programmingCompilers / LLVM

AI Infra: LLM inference / trainingPyTorchONNX RuntimeTritonMoE kernels

Tooling: gdb / DDTCMakePerf / VTuneCI (CTest/PyTest)