Robot Learning · NLP · AI-native Products · Operational Intelligence
I turn field knowledge, operational constraints, and human context into AI systems that work beyond the lab. My work spans autonomous off-road robotics, AI evaluation, logistics intelligence, cybersecurity, and product engineering.
- 90% lower detention costs and 98.7% inventory accuracy through data-driven logistics optimization at Samsung SDS America.
- Built and operated a marketing chatbot that served 25M users, alongside a trend-analysis service for data-driven marketing.
- Published research in IEEE Access, NAACL Findings 2025, and KCI journals; co-authored the book Data Mining Learned with Python.
- Led a fully IITP-sponsored visiting research project at Purdue University on privacy-risk evaluation for drone imagery.
- Defended mission-critical military networks as a CERT security engineer at the R.O.K. Navy Cyber Operations Center.
Dalton, Georgia · Feb 2025 – Feb 2026
- Reduced detention costs by 90% and raised inventory-process accuracy to 98.7%.
- Engineered warehouse-layout simulations and integrated SAP with Cello WMS to improve outbound visibility.
- Converted implicit field knowledge and SOP–reality gaps into better operational system logic.
Kunsan, South Korea · Mar 2023 – Feb 2026
- Conducted research in autonomous off-road driving, sentiment intelligence, and automated essay evaluation.
- Contributed to 2 KCI papers, 1 SCI paper, and NAACL Findings 2025.
- Participated in two industry–academic cooperation projects and co-authored a data-mining textbook.
- Designed experiments for Representation-to-Creativity (R2C) and presented the work as the paper's second author at NAACL 2025.
West Lafayette, Indiana · Mar 2024 – Jun 2024
- Led a fully IITP-sponsored research team in collaboration with Purdue's Computer and Information Technology department.
- Developed methods for evaluating privacy infringement in drone-captured imagery.
Daejeon, South Korea · Jul 2021 – Jan 2023
- Monitored firewall, IDS/IPS, and SIEM feeds as part of the Navy CERT.
- Built Splunk queries and a web service to visualize real-time intrusion attempts and threat patterns.
- Detected and defended against Log4j exploitation without a major breach.
- Representation-to-Creativity (R2C): Automated Holistic Scoring Model for Essay Creativity · Findings of NAACL 2025, pp. 5272–5290.
- Efficient Sentiment-Aware Web Crawling Methods for Constructing Sentiment Dictionary · IEEE Access 9, 2021, pp. 161208–161223. DOI
- Subtitle Automatic Extraction System for Short-form Contents · Journal of Korean Institute of Information Technology 19(6), 2021, pp. 29–37.
- Creative Essay Training Data Labeling System for Learning AI Models · Journal of Korean Institute of Information Technology 21(10), 2023, pp. 69–82.
- Gold Award, Undergraduate Thesis Competition, KIIT Conference, 2023
- Chief of Naval Operations Prize, OSAM Hackathon, 2022
- 3rd Prize, Youth Metaverse Content Development Competition, 2023
Generated from the open-source OpenUsage official local API. Figures cover the latest 30 days and are API-equivalent estimates, not actual subscription spending. Only aggregate metrics are published; prompts, source code, project names, and local paths remain private.
Building systems that make human context executable.