Looking for a job

jihun rho

@jihunrho

CUDA & GPU performance engineer at Enclony who optimized industrial inspection image processing for up to 210,000 tablets per hour.

South Korea

Message

What I'm looking for

I'm open to remote contract work.

At Enclony, I developed and optimized production CUDA/C++ image-processing components for a pharmaceutical inspection pipeline handling up to 210,000 tablets per hour. I tuned GPU kernels, memory usage, and thread/block configurations for performance-sensitive workloads.

I used NVIDIA Nsight Systems and Nsight Compute to investigate CPU-GPU execution flow, synchronization, latency, and kernel performance. I also optimized gradient-based feature extraction and template matching for engraving, imprint verification, and defect detection.

At Sejong University’s Intelligent Vehicle Recognition Systems Lab, I built computer-vision training and evaluation workflows and worked on image datasets, camera calibration, and 3D image projection. At INA Oriental Motor, I integrated robotics and 2D/3D vision systems, and in my current AI Orchestration Program work, I develop AI applications using FastAPI, PostgreSQL, RAG, and LangGraph.

Experience

Work history, roles, and key accomplishments

Current

AI Orchestration Program

Current

SK Networks Family × Encore AI Campus

Jan 2026 - Present (9 months)

Developed and deployed AI applications using FastAPI, PostgreSQL, RAG, and LangGraph. Designed an AI evaluation workflow with 12 parallel evaluation branches and backend integration.

fastAPI PostgreSQL RAG LangGraph AI Evaluation Backend Development

Research Engineer, CUDA / GPU Optimization

Enclony

Mar 2021 - Oct 2025 (4 years 7 months)

Developed and optimized production CUDA/C++ image-processing components for a pharmaceutical inspection pipeline handling up to 210,000 tablets/hour. Tuned CUDA kernels, GPU memory usage, and block/thread configurations for performance-sensitive imaging workloads.

CUDA GPU Optimization Performance Profiling Nsight Systems NVIDIA Nsight OpenCV C++Multithreading Image Processing Template Matching

Researcher

Sejong University, Intelligent Vehicle Recognition Systems L

Nov 2019 - Oct 2020 (11 months)

Built computer-vision training and evaluation workflows with YOLOv3/v4, MobileNet-SSD, LeNet, and SVM. Prepared and curated large image datasets and annotations, and worked on camera calibration and 3D image projection.

YOLOv3 YOLOv4 MobileNetSSD LeNet SVM Computer Vision Dataset Curation Camera Calibration 3D Projection

Software Engineer, Robotics & Vision

INA Oriental Motor

Jan 2018 - Oct 2019 (1 year 9 months)

Integrated autonomous mobile robots, six-axis collaborative robots, and 2D/3D vision systems using Python, C, REST APIs, TCP/IP, Modbus, and digital I/O. Implemented robot-motion and gripper monitoring/control workflows and supported deep-learning-based 3D object recognition and bin picking.