Experience

• 

  Seoul national university of science and technology

  Mar 2017 - Feb 2020

  Student assistant

  I was translating announcements from Korean into English and helping to find communication between the administration of dormitory office and foreign students.

• 

  Brix academy

  Mar 2019 - May 2019

  Tutor

  I was preparing a student for AP Computer Science Principle exam.

• 

  Im2g

  Aug 2019 - Oct 2019

  Data engineer

  I was sorting, clearing and visualizing data. Created MySQL database and moved all Excel files to the new database, created manager pages to view statistical reports as well as inserting new data. Finally, put the web server and database to AWS server.

• 

  Data mining lab (snu)

  Feb 2020 - Jan 2022

  Artificial intelligence researcher

  I was working on two projects Accurate Company-Related News Extraction and DRAM log data processing.- Accurate Company-Related News Extraction Collected news from Naver.com and used it to train an Attention model that can extract company names from Korean news articles. The research results were published at the KIISE, and a detailed version was used as the graduation thesis.- DRAM log data processing for Samsung I've built a model that can process a long sequence of data (100k commands) and represent it into vectors of smaller size for further classification. The model can detect unknown data similar to "open set recognition" and "anomaly detection models". Original data was in several Terrabite scale log data. Show less

• 

  Deeping source inc.

  Mar 2022 - now

  - Auto-Labeling Pipeline Integration - Collaborated with cross-functional teams to integrate an auto-labeling pipeline into the company’s workflow. - Utilized CI/CD pipelines, Terraform, and Atlantis for efficient deployment and management.- Prototyping and Productionizing Model Inference Pipelines - Developed a prototype of a model inference pipeline using N8N, enabling quick testing and iteration. - Transitioned the prototype to a scalable production version: • Deployed as AWS Lambda and ECS services. • Configured comprehensive infrastructure and networking on AWS using Terraform (Infrastructure as Code). • Automated deployment processes to streamline production readiness.- Synthetic Data Generation - Leveraged Unity to generate extensive synthetic datasets for model pre-training in new domains. - Improved model training efficiency by providing high-quality data for initial domain adaptation. Show less As an MLOps Engineer at Deeping Source, I specialize in developing advanced ML model pipelines and enhancing customer insights through innovative data analysis.- Model development I have developed a GAZE model that can detect the direction of a pedestrian's gaze from a CCTV video. I further developed a model that can analyze a person's attention towards some specific objects using the GAZE model.- Auto-labeling pipeline I have developed a pipeline that can be used for automatic labeling of data. The pipeline is created on top of AWS infrastructure and utilizes such services as Terraform, AWS S3, Lambda, Step Functions, Batch, and others. The pipeline is built with modular architecture where each module can be easily replaced with another one. Show less

  • Senior Machine Learning Engineer

    Jan 2024 - now

  • Machine Learning Engineer

    Mar 2022 - Dec 2023