Dongook Son
- +821026688138
- dongook.son@yonsei.ac.kr
- Blog
- Seoul, Korea
I envision a world where making data-driven decisions is hassle-free and instantaneous. Despite recent technological advancements in the field of AI and Data Science, I have witnessed that implementing data-driven decision/policy making is still not viable for most small organizations. I hope to dismantle this barrier by constantly listening and providing useful tools to the decision makers.
Work Experience
CEO
Currently creating variety of data crawling systems and visualization modules for local government congressmen.
Big Data(Spatial) Engineer
Currently working on transparent geodatabase(PostGIS) creation on AWS with automated data collection using govermental OpenAPIs for air quality and health research team led by Dr. Sunyoung Kim. Containerized and debugged existing multiprocessing geostatistics calculation software(R). Other researchers were able to gain insights and comfirm hypothesis from statistics dervied from this software. Introduced and implemented co-research environment with Notion. Modified numerous data standardization processes.
Data Engineer/Statistician
Created multiple data extracting software to accumulate mass amounts(50GBs xml, 400GBs jpg data) of patient records/ocular scans for data analysis/AI research for Opthalmology research led by Dr. Jongchul Han. Experienced app containerization and cross-platform builds due to security issues for on-premise networks while working remotely. Built fundus scan image preprocessing software using Go. Completed research and co-authored Border Tissue Morphology is Associated with the Pattern of Visual Field(acceptance pending). Converted and upgraded a mid-performance workstation(CUDA enabled) as a virtual research lab with a dedicated NAS for instant sharing and usage of preprocessed data. Currently conducting research on following projects.
- Longitudinal analysis of health screening data, intraocular pressure and glaucoma risk.
- Multiple instance learning using Resnet50 architecture for unintrusive health pre-screening through imagery data(OCT scans).
HPC System Administrator/Server Developer
Funded by BK21 Educational Research Group for Big data and Data Science, I was able to design/configure/manage/operate High Performance Data Science servers(High RAM, vRAM cloud instances) grateful to academic advisor Prof. Sangun Park and Prof. Ikhoon Jin. Researched numerous HPC systems like Nero, Niagara, Great Lakes and their operations framework such as SLURM, Kubernetes. Implemented optimal JupyterHub, RStudio Server service for all Statistics and Data Science graduate researchers.
HPC System Administrator/Server Developer
Configured and managed High Performance Computing server for supply-chain finance research team led by Prof. Junkyeong Auh. Introduced and educated researchers on DevOps concepts such as app containerization, version control and test-driven-development which solved numerous works in my machine problems. Implemented Graph database modeling(Neo4j) from relational supply-chain datasets to apply network analysis.
Data Scientist/Developer
Data development for accurate prediction of court auction product prices. Worked with several OpenAPIs and structured various relational data models. Built and modified custom web crawler with Selenium framework(Python) for Court Auction sites.
Imagery Intelligence(GIS) Analyst/Interpreter
Honorable discharge after 20 months of service. During this time I successfully completed 3 weeks of Airborne training(#826) at Special Warfare School with two fractured ribs. Performed numerous GIS analysis for indisclosed locations and operated as interpreter on combined training periods. On completion of a 5 week interpreter training at Special Warfare Command, conducted 3+ combined special ops training with U.S. counterparts including 1st Battalion of 1st Special Forces Group.