Experience

LBox Co., Ltd.

• Tech Lead

  • Jul 2022 - Present

  - Engineering/Data/AI 팀 빌딩 : 3명 → 약 13명 ㄴ 다이렉트 소싱, 채용 프로세스 정리 및 엔지니어링 팀의 개발문화 형성 (스터디, 세미나 진행 등)- 내부 시스템 개편 진행 ㄴ 팀 확장을 위한 Angular → React 전환 결정 ㄴ 기존 데이터 중심의 백엔드 시스템을 도메인 중심으로 리펙토링- AWS ECS 기반으로 인프라 구성 및 운영- 모니터링 시스템 도입 및 전환 (Datadog)

• Machine Learning Engineer

  • Apr 2021 - Jul 2022

  - Implemented data pipeline with Airflow and improved overall search system (e.g. query completion, suggestion, optimize Elasticsearch query performance, and so on)- Implemented a deep learning-based dense retrieval system for searching similar cases (10% improvement based on Retrieval Top 1 Accuracy)- Implemented Model serving pipeline with BentoML.- Pre-trained Legal-domain language model as the Core of the NLP engine.- Implemented monitoring system (Elastic APM, Sentry, Grafana, Prometheus)- Set up and operated BI tools for data-driven communication Show less



NAVER Corp

• South Korea
• Technology, Information and Internet
• 700 & Above Employee

• Technical Lead

  • Aug 2020 - Mar 2021

  - Led the project that automates the entire process of E-commerce collections ('기획전' in Korean)

• Project Manager

  • Sep 2019 - Jul 2020

  - Manage the delivery of AiCall service for the Outback steakhouse (AI system for accomplishing real-world tasks over the phone instead of the restaurant owner)

• Machine Learning Engineer & Project Manager, Clova AI

  • Apr 2018 - Jun 2020

  - Led the project that develops a Large-Scale Language Model pipeline for Korean and Japanese (e.g. BERT, GPT-2, RoBERTa)- When our team submitted, achieved first place with F1 94.75 on KorQuADv1, F1 83.54 on KorQuADv2 (Korean version of SQuAD) Test Set- Gave a talk at Deview 2019 (https://deview.kr/2019/schedule/291)- As the author of the open-source Clova Language Framework (https://github.com/naver/claf), I designed and implemented the project and am maintaining this project



Konolabs Inc.

• Software Engineer, Machine Learning Engineer

  • Jul 2016 - Mar 2018

  - Led the Kono Slack Bot project that schedules a team meeting as easy as possible - Implemented Email Bot pipeline with AWS and NLP Server (Django + TF Serving) - Researched NLP based on Deep Learning and mainly worked to expand coverage of temporal information extraction based on Rule-based - Led the Kono Slack Bot project that schedules a team meeting as easy as possible - Implemented Email Bot pipeline with AWS and NLP Server (Django + TF Serving) - Researched NLP based on Deep Learning and mainly worked to expand coverage of temporal information extraction based on Rule-based


Lunit Inc.
• Software Engineer Intern

  • Jan 2016 - Jun 2016

Minz Communication Inc.
• Software Engineer Intern

  • Jun 2014 - Oct 2014