If you are intereted in me, you can get more detail about me by visiting my Portfolio page for detail!!
👉 Notion Portfolio 👈
I’m a software engineer who is curious about every new things. And “Keep Blitz and Be Simple”, “Keep It Simple Stupid” is my motto. I always try to be simple. 😊
🌥 SOFTWARE ENGINEER
Proficient in end-to-end planning and operation of MLOps platform as a Software Engineer
Skilled DevOps Engineer well-versed in Cloud Native technologies
Experienced System Engineer with expertise in operating large-scale HPC clusters, managing over 2000 GPUs
INTERESTS
🏢 WORK EXPERIENCES
Toss Bank
Machine Learning Software Engineer, Machine Learning Platform Team
Oct 2023 ~ Present
- Built Machine Learning Platform, contributing to flexible model development and deployment.
- Pioneered a high-performance Online Feature Store Platform leveraging ScyllaDB (Cassandra-compatible, high-performance NoSQL DB), achieving performance improvement in write throughput and read latency. Developed a Spring-based feature store API server to ensure low-latency feature access at scale.
- Engineered and sustained a large-scale Airflow batch system managing 2K+ workflows across 7+ clusters, achieving 99.9% system uptime in the critical banking environment while optimizing DAG management to reduce batch processing time by 20%.
- Contributed the operation of an AWS-based research network, boosting productivity by reducing research time by 50% with fortified data security measures.
- Built a caching system using MongoDB, reducing errors by 60% within the credit scoring system, directly improving loan execution and driving revenue growth.
- Migrated a data ingestion pipeline processing 9M+ daily records from Elasticsearch to MongoDB, enhancing data security and compliance, which contributed to reliable and scalable data handling.
- Developed Python-based Credit Strategy System that cut deployment time by 50%, reduced costs and errors through automation, and enhanced credit decisions using Machine Learning models
Kakao Enterprise
Software Engineer, Machine Learning Platform Part
Mar 2022 ~ Aug 2023
Built large-scale MLOps platform that can manage the entire process from “Traning” to “Inference” and production model deployment.
- Developed and maintained the Kakao i Machine Learning MLOps platform leveraging Golang and Python. Managed over 2,000 GPUs(NVIDIA A100, V100, T4) across 20+ Kubernetes clusters using Kubespray. Ensured smooth platform operation for approximately 200+ active users, achieving 99.9% system uptime.
- Successfully operated a stable HPC(High-Performance Computing) cluster basede on Infiniband network. Also established and maintained stable infrastructure for high-performance computing projects in public environments.
- Planned and executed the successful relocation of over 300 GPU, CPU devices for a major power facility relocation with 99.99% system uptime.
Developed MSA components in the backend (API Server), client (CLI, SDK) and infra element development of inference function for MLOps platform. And also built CI/CD system for platform over Kubernetes.
- Developed managed Jupyter Notebook functionality for the MLOps platform by integrating open-source components such as JupyterHub and Enterprise Gateway, enabling users to launch notebooks with dynamically changeable kernel instance types.
- Implemented Model Artifact resource functionality for the MLOps platform, including CRUD APIs, metadata support for model versioning, and tagging capabilities to enhance model traceability and management.
- Established a unified API server error code policy by aligning multiple backend teams, facilitating clearer communication across engineering, frontend, and planning teams.
- Led GPU migration and service continuity planning for the Kakao i Machine Learning platform, including the relocation of 200+ high-performance GPU devices. Collaborated cross-functionally to ensure seamless integration into Kubernetes clusters and maintained inference service stability during NAS and infrastructure transitions.
Software System Engineer, AI Lab Voice System Part
Jul 2020 ~ Feb 2022
Collaborated within the AI Lab, building infrastructure and tools to support voice-related research in an organization of 100+ AI researchers.
- Designed and operationalized the Voice Transcription Tool with React and Django(Python), enabling 10+ transcribers to process 3K+ voice data daily, significantly enhancing transcription productivity and addressing domain drift issues in STT(Speech-to-Text).
- Operated a Kakao i voice data log management system to handle 3M+ daily data logs from 100k+ users. Ensured data security by isolating sensitive voice data in secure zones, and streamlined data refinement to support transcription workflows and researcher access.
- Built and operated the Kakao i Cloud STT service, enabling 20+ custom user language models on Kubernetes through a Python-based controller, resulting in efficient deployment and stable operation.
- Developed real-time STT demo services with WebSocket streaming and file upload support to showcase Kakao i’s speech recognition features.
🛠 SKILLS
Language
- Python
- Go
- Javascript
Cloud Native
- Kubernetes
- Docker
- ArgoCD
- Kubeflow
DevOps
- ArgoCD
- Github Action
- Jenkins
- Helm
Keywords
- Microservice architecture
- Linux/System programming
- Machine Learning
📜 CERTIFICATIONS
AWS Certified Cloud Practitioner (CLF)
Issued Apr 2023 · Expires Apr 2026
MLOps Engineering on AWSMLOps Engineering on AWS
Issued Apr 2022 · Expires Apr 2025
🎓 EDUCATION
Kyungpook National University
2015.03 ~ 2021.02.
- Bachelor of Computer Science
- GPA : 4.10 / 4.50
Darmstadt Univ of Applied Sciences, Germany
Spring Semester 2020
- International Studies Program