--- Achievements
• Measured optimal liquidation time and costs for the given liquidity policies (e.g. Liquidation size 100% for max. 30 day trading periods, and portfolio ES limit)
• Estimated and compared (prior/posterior) VaR/ES on the distribution of financial costs given the department selected in the Bayesian network.
--- Responsibilities
• Implemented an application to measure portfolio market liquidity risk employing Power-Law and stair step MSDCs, and time average VaR/ES (Matlab, Java, MySQL)
• Developed an application to quantify operational risk using Bayesian belief network

[UN peacekeeping operation events coding 시스템 개발]
--- Responsibilities
• Extend the TABARI software (http://eventdata.psu.edu/software.dir/tabari.html) to automate peacekeeping operation events coding (C++)
• Creating an web interface that allow users to code UN peacekeeping operation reports and to maintain actors, events dictionaries (Java Vaadin, PostgreSQL)
--- Achievements
• Improved the quality of coding (over 60% recognition rate in comparison with the human coders)

[ExpCapture and anon-ID`s 시스템 개발]
--- Responsibilities
• Setting up a “Dataverse” (open-source software developed at Harvard), an interface for depositing and linking data, for social science lab experiments (Glassfish, PostgreSQL)
• Developed a data feeder (ExpCapture system) that upload Ztree data to Dataverse in real time (Java NIO)
• Implemented Experiment Participants Identification key generator (anon-ID`s system) (JSP, MySQL)

- Java, mySQL 운영

- Bayesian 네트워크 (조건) 변수로 선택된 부서별 P&L 99% VaR, 97.5% ES 산출 (Matlab, MySQL, Java)
- Stair step MSDC 이용한 포트폴리오 Liquidation time, cost 산출 (Liquidation size: 100%, Max. liquidation periods: 30일)

--- Data Analyst, POSCO (Client), Seoul, Korea (1/2014 ~ 6/2014)
Achievements
• Improved prediction for nickel price (both direction, size) with market sentiment indicator (60%)

Responsibilities
• Implemented Nickel price forecasting system to use on buy decision making (Java, PostgreSQL, SAS)
• Devised patterns of text that indicate future nickel price movement and trend of risk factors
• Coded economic events from Reuters news and quantified the impact of events using a word frequency distribution

--- Senior Programmer, Samsung Electronics (Client), Suwon, Korea (9/2013 ~ 12/2013)
Achievements
• Reduced the time of startup sensing process (2 hour frequency)

Responsibilities
• Implemented startup sensing system to use on merge and acquisition (M&A) deals
• Created batch jobs to extract, transform and load the M&A transactions data by Standard and Poor’s Capital IQ into the local DB (Java, MSSQL, SAS)