Gyuhyeok
Park박규혁
Curiosity takes different forms.
I study Quantitative Risk Management at Yonsei University and build tools around the problems I encounter. I create independently, seek feedback, and refine the work.
Education
Yonsei University · Quantitative Risk Management
Statistics, econometrics, financial data analysis and financial engineering
Experience
Assistant · Disability Support Center, Yonsei University
Administrative assistant · Graduate School of Social Welfare
Student work-study · disability support
Community
QRM student council · Participation & Communication Director
Academic information · @uic.qrm.info
KUSA · YCC · enVision
Recognition
Grand Prize · AI Ideathon, Yonsei University Library
ADsP · Advanced Data Analytics Semi-Professional
From administrative work to accessible learning
My experience includes student disability-support work in 2025, administrative work at the Graduate School of Social Welfare from March to August 2026, and an assistant role at the Disability Support Center from September 2026. Producing learning materials for visually impaired students led to Hun-maeng-jeong-eum: a service with student–course profiles and separate AI production and review roles.
I have also recorded 208 hours of volunteer work. My interest in accessibility includes helping people participate as active contributors, with autonomy and self-efficacy, rather than defining them only through the support they receive.
Information, communication, and coordination
Since January 2026, I have operated academic-information content at @uic.qrm.info. I use AI-assisted workflows with human oversight and direct review at important steps to balance accuracy with the time available for organizing and publishing information.
As Participation & Communication Director in the QRM student council, my experience includes meetings, project coordination, and events up to approximately KRW 2 million and 40 participants. I also participate in KUSA, YCC, and enVision.
Other experiments with AI
I built a residual-caffeine calculator and separated information-gathering agents from agents checking validity and reliability. I have also built an MCP connection for subagent workflows across different models, exploring how to use a limited AI budget effectively. These experiments connect research, verification, and resource allocation.
Languages and working style
Korean is my native language. I am comfortable reading and writing in English and discussing subjects in my field, while everyday spoken English remains an area I am developing. I work across independent building, feedback-driven iteration, and team coordination.