Causal inference · Experimentation · Marketplaces

Economist by training.Data scientist in practice.Program manager in operations.

I'm Youngmin Ju, a USC economics PhD. I use causal inference to settle what actually worked, and I've taken it from research to ML products to operations at Coupang.

See my experience LinkedIn Email

Ph.D. Economics: University of Southern California
Sr. Manager, Program Management: Coupang
Co-founder: AIEconLab, AI economics research
Published: Oxford Bulletin of Economics & Statistics

Fig. 1Career path, 2004–2026

2026Operations
Coupang
Sr. Manager, Program Management
2025Applied ML
VWS
AI lead, NPL recovery
2023Research
AIEconLab
Co-founder · AI + economics posts
2022Applied ML
Datacrunch Global
Forecasting & data pipelines
2021Applied ML
Decode Data
Founder, applied ML
2017Research
Oxford Bulletin of Economics & Statistics
Paper on causal identification
2015Education
USC
Ph.D. Economics, 2015–2021
2010Research
Korea Army Academy
Asst. professor of economics
2009Research
Economic Researcher
Korea Univ. · Hyundai Mobis case
2008Education
Korea University
M.A. Economics, 2008–2010
2004Education
KAIST
B.S. Mathematical Sciences, 2004–2008

Fig. 1. Degrees and roles from 2004 to today, by the kind of work they were. Dashed: AIEconLab, co-founded in 2023 and run alongside.

01Evidence

Three kinds of work, one method.

Causal inference is the thread: working out what caused an outcome before anyone acts on it.

Research

A published method, and AI economics in public

A control-function method for partly ordered treatments, in the Oxford Bulletin of Economics and Statistics. Through AIEconLab, which I co-founded, I post on AI and economics — most recently a working paper on AI adoption in Korea.

Read the latest piece

Applied ML

From model to measured lift

At VWS I designed an LLM + behavioral-ML pipeline for loan recovery and judged channels by DID and uplift, not response rates. Recovery rates rose 1.2–1.3x and collection cost fell 15%.

See the details

Operations

Analysis the business runs on

I lead Business Operations for Rocket Growth Consumables at Coupang, building fee-scenario analysis, seller-level P&L simulators and the team's shared analytics.

See my experience

02Writing

Recent research & writing

All posts

03Projects

Selected projects

All projects

Netflix Top 10 Analytics Dashboard

Interactive dashboard built on Netflix's published weekly Top 10 data, with real hypothesis tests and a logistic regression model trained to predict which titles hold their chart position.

Data Science

Dashboard

Hypothesis Testing

Military Rank Premium Analysis

Statistical methodology to analyze the effects of military ranks on wages, published in Oxford Bulletin of Economics and Statistics.

Econometrics

Causal Inference

Control Function

Analyzing Wage Determinants with Advanced Causal Inference

Comprehensive analysis of wage determinants using OLS, Random Forest, Double Machine Learning, and Causal Forest to identify causal effects and heterogeneous treatment effects.

Causal Inference

Double Machine Learning

Causal Forest

Hiring for causal inference, experimentation or marketplace analytics?

I'm glad to talk.

Email me LinkedIn