Dongwon Kang

• Develop based on Serverless Architecture (kubernetes, AWS Lambda) and deployment (terraform)
• Backup service development & maintenance, based on Golang
  • Service improvement and development
  • Code quality improvement: Test Coverage, Refactoring, Peer review
  • Documentation for design/development
• Big-scale scheduler
  • Schedule 2~3 billion tasks w/ requirements
• Monitoring tool and tracing in MSA

• Infra based on Terraform + AWS
  • Designed for future extension & monitoring capability
  • Used techniques: CW + ECS + ElasticCache + MySQL + MongoDB + SQS
• Backend design and development
  • Based on Node.js + Go + protobuf
  • Developed user tracking & position share service
• CI/CD
  • QA automation & deployment through Github Actions
• Management
  • Blocking prevention: Do short but often meeting & sprint
  • Writing document and encouraging

• Managing issues
  • Do regular meeting to manage blocking items
  • Managing techinical debt for long term issues
• Improving individual skills
  • Study Books: Refactoring 2, Clean code
• Improve SWE culture
  • Code review to improve readability and test coverage.
  • Encourage documentation to the members
• Improve effciency
  • Reduce unnecessary meetings and issues
  • Documentation and tutorial for New-hire

RDBMS engine develop & maintenance w/ C/C++

• Optimization
  • Minimizing lock to better performance / removing deadlock.
  • Improve caches (partition, global cache, etc.)
• Feature development
  • PL/SQL engine development
    • Nested Type
    • Dblink protocol
    • PL/SQL VM interpreter optimizing
    • PL/SQL compiled mode design
  • Parser design & development
    • Based on Bison/Flex (DDL, PLSQL, SQL Parser)
    • Custom Parser design to reduce grammar conflict
  • PL/SQL Package development
    • Message Queue based on PL/SQL
    • Debugger based on PL/SQL
  • Gateway
  • Job module
  • DB monitoring process
• Improve testing
  • CI pipeline management (add test)
  • Introduce unit test
• Improve development culture
  • Peer code review
  • Write and encourage documentation
  • Add test cases
  • Create docker container for constant development/testing
• Output
  • Removed most of the severe bugs of our module.
  • Contribute to company for developing important features, about 3x Revenue compared to 18y.

• Various side project on Github
• Development activity: UPnL
• Security activity: Guardian
• Some events: SW Maestro, Microsoft App Hackathon 등

• Streaming platform of individual (Android dev.)
• Game controller platform (Android, server, demo game dev.)

• Analyzing mRNA data
• Model design and develop based on Tensorflow
• Written 4 Paper (one main author)

• Model development which find out stress label and biomarkers from large matrix data (time-series mRNA).
• Reduced dimension by pre-processing matrix data (Calculate DEG p-value and one-hot encoding)
• Added group effect to avoid overfitting
• Used CMCL loss for multiple label selection

• In silico experiment delivered as Web application
• Front: jquery + bootstrap + cytoscape.js
• Backend: Flask + Rscript for data pipeline

Venn-diaNet: venn diagram based network propagation analysis framework for comparing multiple biological experiments

• Improved venn diagram rendering method for biomarkers
• Venn diagram and node element positioning using Edge weight, which makes capable for N-group diagram
• Optimize: performance improvement using threshold for selective rendering.

HTRgene: integrating multiple heterogeneous time-series data to investigate cold and heat stress response signaling genes in Arabidopsis

• New DEG Algorithm: Using clustering, find order of biomarker and it’s pvalue.
• Performance Proof: Compared to other DEG algorithm, HTRGene had found remarkable genes other tools hadn’t found with Arabidopsis.

• Game Engine by my own
  • Open source
  • Based on GLFW
  • Performance optimized: multicore resource loader, task scheduler for loading
  • Lag/Delay minimized: dynamic vsync, async event-handler
  • Mixer implementation with Sound effector
• Multimedia data processing
  • Write decoding for various audio/video/image
  • Support for resource of other game (dxa)
• Game script engine development
  • External script support (lua)
  • Game script runtime (lr2script)
  • Be-music script library engine
• CI
  • Used googletest

• Audio data processing
  • Audio decode/encode, mixing development with pitch / speed change
• CI: travis CI for QA automation

• Game record managing
  • frontend: jQuery + bootstrap
  • backend: django + mariadb(ORM)
• Recommendation service
  • Analyze user data using jupyter notebook
  • Learn user data through Logistic regression & cross entropy
• CD based on shell script.

• client: Android, server: C#

• server: python, client: Android
• Motion recognition using machine learning for high accuracy

• Based on libGDX engine
  • High compatibility (Windows, Mac, Android)
  • Small time/cost for development

• Text novel viewer for UMPC
• Many features: bookmark, automatically open last closed file, etc.
• Smooth pagination animation
• Performance optimization: DirectX(GPU) + font cache for better UX