Portrait of Thanh Loi Hoang

Thanh Loi Hoang

Graduate Research Engineer · Software Engineer · Seoul, South Korea

I work on the GPU layer underneath Kubernetes AI workloads — sharing, scheduling, and observability — at Cloud Native x AI Laboratory, Soongsil University. I learn a system by standing it up and watching it, and I'm comfortable at the layer where the abstractions leak.

News

  • 2026-08Co-authored "Synchronous Distributed Training with Runtime-Adaptive Mechanisms in Hybrid Cloud Environments" in IEEE Access.
  • 2026-07Joined HyeraEdu as Software Engineer & Technical Partner, designing and shipping a multi-service web platform.
  • 2026-06First-authored "A Design for Continuous Low-Overhead CUDA-API Observability in Kubernetes GPU-Sharing Systems" at the KICS Summer Conference.
  • 2025-06First-authored "A Design of Cloud-based Personal Machine Learning Workspace" at the KICS Summer Conference.
  • 2025-03First-authored "A Comparison of Two SDN Controllers Integration in OpenStack Environment" at ICISN.

Featured projects

All projects →

ML Platform

2025 — Present

Cloud Native x AI Laboratory

Kubernetes-based platform for machine learning workloads, supporting interactive development and training. Covers GPU sharing across heterogeneous hardware, scheduling and admission control for compute-intensive jobs, internal tools that make common workflows easier to use, and the core infrastructure for images, data, and artifacts. A monitoring layer underneath correlates device-level and process-level GPU activity with the workloads that produced it.

OpenShift Deployment

2024 — 2025

FPT IS

End-to-end Red Hat OpenShift cluster deployments for banking environments, from VM preparation and prerequisite configuration through cluster installation and platform component setup.

Publications

Research →
  1. Synchronous Distributed Training with Runtime-Adaptive Mechanisms in Hybrid Cloud Environments

    T. A. Vuong, T. L. Hoang, H. Le, Y. H. Kim

    IEEE Access · 2026

  2. A Design for Continuous Low-Overhead CUDA-API Observability in Kubernetes GPU-Sharing Systems

    T. L. Hoang, Y. H. Kim

    KICS Summer Conference · 2026

  3. A Design of Cloud-based Personal Machine Learning Workspace

    T. L. Hoang, T. N. D. Thong, Y. H. Kim

    KICS Summer Conference · 2025

Photos

Conferences, campuses, and places in between.

  • KubeCon + CloudNativeCon Japan, Yokohama
    KubeCon + CloudNativeCon Japan, Yokohama · Jul 2026
  • KICS Summer Conference, South Korea
    KICS Summer Conference, South Korea · Jun 2026
  • Busan Station, Busan, South Korea
    Busan Station, Busan, South Korea · Aug 2025
  • Soongsil University, Seoul, South Korea
    Soongsil University, Seoul, South Korea · Mar 2025
  • Nui Coc Lake, Thai Nguyen, Viet Nam
    Nui Coc Lake, Thai Nguyen, Viet Nam · Oct 2024
  • VNU University of Engineering and Technology, Hanoi
    VNU University of Engineering and Technology, Hanoi · Jul 2024
  • REV-ECIT 2023 national conference, Hanoi
    REV-ECIT 2023 national conference, Hanoi · Dec 2023