2025–Present
Co-Founder, CEO & Chief AI Scientist · insightea
Seoul, South Korea · Leads AI research and platform architecture for personalized news, recommendation, and agentic reasoning.
Woojin (Jay) Park
Co-Founder, CEO & Chief AI Scientist at insightea
I lead AI research and platform architecture for personalized news, recommendation, and agentic reasoning at insightea. I am also a fourth-year Ph.D. candidate in Industrial Artificial Intelligence at Korea University, advised by Prof. Taesu Cheong in the SaVAnNA Lab.
My work spans recommender systems, user modeling, large language models (LLMs), algorithmic auditing, and reliable AI. I connect research in measurement and evaluation with production machine-learning systems in news, healthcare, and large-scale consumer platforms.
I received an M.S. in Data Science from Carnegie Mellon University after graduate study and research at Syracuse University in the Science of Science & Computational Discovery Lab, led by Prof. Daniel Acuna. I bring more than a decade of industry experience in data science, applied AI research, and technical leadership, including roles at Hyundai & Kia Motors Group, IQVIA, Coupang, and Catchtable.
I am exploring opportunities in applied science, data science, and AI research, including technical leadership roles and postdoctoral positions, with a flexible start date in the first half of 2027. I would be happy to connect and discuss potential opportunities.
Recent News
- Released the public research artifact for the PEC news-recommender audit.
- Released the PEC Gap preprint on arXiv (under review at AAAI ICWSM 2027).
- Was granted a Korean patent for a user-controlled equalizer interface for real-time personalized news recommendation.
- insightea joined NVIDIA Inception to advance GPU-accelerated deep-tech AI research and development.
- Opened Google Cloud Day Seoul's AI Playground with a talk on Catchtable's production AI and personalization.
Experience
2024–Present
Doctoral Researcher · Korea University
Seoul, South Korea · Conducting research on recommender systems, large language models, algorithmic auditing, and reliable AI in the SaVAnNA Lab.
2024–2025
Staff Data Scientist · Catchtable
Seoul, South Korea · Led recommendation, personalization, ranking, advertising models, and production ML architecture.
2023–2024
Senior Data Scientist · Coupang
Seoul, South Korea · Led supply-chain forecasting and optimization across large-scale fulfillment operations.
2022–2023
Chief Technology Officer · Coffeechat
Seoul, South Korea · Built and led the AI and engineering organization, recommendation systems, and production MLOps platform.
2021–2022
Senior AI Scientist · IQVIA
Cambridge, Massachusetts, USA · Machine Learning Analytics Center of Excellence; led cross-border machine-learning programs for clinical-trial decision intelligence.
2020
Graduate Data Scientist Intern · Neolth
Walnut Creek, California, USA · Developed NLP and behavioral-anomaly models for mental-health risk detection.
2020
Graduate Data Analytics Consultant Intern · KPMG
Pittsburgh, Pennsylvania, USA · Built customer segmentation and predictive models for acquisition and retention.
2019
Data Science Researcher · Carnegie Mellon University
Pittsburgh, Pennsylvania, USA · Conducted landslide-risk research and delivered recommendations to North Huntingdon Township.
2018–2019
Data Science Researcher · Syracuse University
Syracuse, New York, USA · Researched graph-based measures and large-scale scholarly data in the SOS+CD Lab.
2014–2018
Data Scientist / Big Data Analyst · Hyundai & Kia Motors Group
Seoul, South Korea · Built forecasting and connected-vehicle analytics across supply chain, manufacturing, and energy.
Selected Publications
When Exposure Is Not Attention: Auditing the Preference-Exposure-Consumption Gap in Personalized News Recommenders
A measurement audit that separates stated preference, logged recommendation exposure, interface pathways, and observed click consumption.
PEC-Fuse: Source-Preserving Evidence Fusion for News Ranking with Partial Recommendation Traces
A source-preserving evidence-fusion framework that keeps profile alignment, retained recommendation traces, and candidate-aware consumption distinct until scoring.
Paper & artifacts coming November 2026
Confidence-Guided Selective LLM Re-Evaluation for Depression-Risk-Related Emotion Classification in Social Media Text
A calibrated two-stage pipeline that routes low-confidence emotion classifications to selective LLM re-evaluation and measures both corrected and introduced errors.
Paper & artifacts coming November 2026
Are Candidate-Weighted History Summaries Enough? Bounded History-Agreement Refinement for Frozen News Rankers
A bounded post-hoc refinement that adds history-wide semantic agreement to frozen news rankers and isolates its value with a capacity-matched control.
Paper & artifacts coming January 2027
Selected Patents
Sole inventor.
-
Real-Time Personalized News Content Recommendation System Having a User-Controlled Equalizer Interface and Method of Operating the Same
Republic of Korea Patent 10-2965849 · Issued May 2026 · [Google Patents]
-
User and Conversation Partner Matching Optimization System Based on a Representative Activity-Log Extraction Method and Machine-Learning Model
Republic of Korea Patent 10-2600307 · Issued November 2023 · [Google Patents]
-
Personalized Recommendation System for Matching Users and Conversation Partners Based on Integrated Machine-Learning Models
Republic of Korea Patent 10-2600305 · Issued November 2023 · [Google Patents]
-
Hybrid Bayesian Inference and Deep Language Model for Partner Recommendation
Korean patent application P20220291-KR0 · Filed February 2023
Invited Talks & Presentations
-
KGU IR Deep-Tech Bridge Demo Day — Deep-Tech AI Presenter; presented insightea's personalization, recommendation, and AI news-intelligence technology. [Coverage]
-
Google Cloud Day Seoul: AI Playground — Opening Speaker; presented Catchtable's production AI and personalization strategy. [Event] [Customer story]
-
Google Leaders Exchange — Invited Speaker, “The Future of AI.”
-
Korea University, Department of Industrial and Management Engineering — Guest lecture, “About Coupang, SCM & AI Tech.”
-
Carnegie Mellon University Policy Innovation Lab — Municipal landslide-risk presentation for North Huntingdon Township, Pennsylvania; featured by PIT-UN as a public-interest technology case. [Case study]
-
Deloitte Consulting Case Challenge, Carnegie Mellon University — Finalist Presenter; awarded second prize.
Education
Korea University
Ph.D. Candidate, Industrial Artificial Intelligence
Expected August 2027Carnegie Mellon University
M.S., Data Science, Business Intelligence & Data Analytics
Syracuse University
M.S., Applied Data Science · Faculty Engagement Scholar
Transferred to Carnegie Mellon UniversityHanyang University
B.S., Information Systems
Contact
For research collaborations, speaking invitations, and professional opportunities, please feel free to reach out.
- Primary email
- woojin_park@korea.ac.kr
- Secondary email
- jay.seizethemoment@gmail.com
- Phone
- +82 10-5681-5788
- Location
- Seoul, South Korea · Open to relocation