Minseo Choi

About Minseo Choi

I am a Computer Science and Computer Engineering student at Johns Hopkins University with a deep focus on Machine Learning Systems, GPU kernels, and compiler-driven performance optimization. I enjoy working at the intersection of systems, deep learning, and high-performance computing, where thoughtful engineering can translate directly into measurable speedups and real-world impact.I’ve built projects ranging from custom Triton GPU kernels for medical image enhancement (achieving up to 10–30× faster runtime) to LLVM-based compiler tooling for IR generation, CFG visualization, and optimization-pass exploration. I also implement modern deep-learning models from scratch to strengthen my understanding of architecture-level design and research-to-code translation.My technical interests revolve around- MLSys: GPU programming, Triton, CUDA fundamentals, memory hierarchy, kernel fusion- Compilers: LLVM/MLIR, IR transformations, JIT pipelines, optimization passes- Deep Learning Engineering: PyTorch, distributed training, performance debugging- Systems & Infrastructure: FastAPI services, Dockerized pipelines, remote dev environmentsI care deeply about building things that are not only correct, but efficient, reproducible, and grounded in systems thinking. Whether it’s writing kernels, profiling IR, reading research papers, or designing reliable experiment workflows, I enjoy the process of turning complexity into clarity.Long-term, I hope to contribute to ML compilers, GPU kernel optimization frameworks, and next-generation inference systems—projects where strong fundamentals and engineering discipline truly matter.If you’re working in MLSys, compilers, GPU programming, or AI infrastructure, I’d love to connect.

Experience

1. Executive Board Member Jhu Chatper

   Acm, Association For Computing Machinery

   Apr 2026 — Present

Education

• The Johns Hopkins University

  Bachelor of Science - BS, Computer Science

• Korea International School, Jeju Campus

  High School Diploma