Kelun Wang
Kelun Wang · Analytics & decision science
Hey there, I’m Kelun 👋 I like turning messy data into clear decisions.
Welcome to my analytics notebook 📓 — a growing collection of experiments, models, and business questions I have enjoyed taking apart.
Currently looking for data, product analytics, and business analytics roles where technical work changes what a team does next.
& targeting
& machine learning
Tableau
decision, not the model.
Selected work
Analysis that ends with an answer.
Each project begins with a business decision and shows the reasoning, model, evidence, and recommendation behind it.
Tuango campaign targeting
Impact: Selective messaging increased realized profit from roughly 112K to 887K RMB.
Logistic regression · ROME · rollout evaluation
A/B testing and peeking risk
Impact: Shows how repeated peeking can inflate false positives from 5% to nearly 20%.
Bootstrap · CLT · hypothesis testing
Card & Krueger replication
Impact: Reproduces a +2.75 FTE difference-in-differences estimate after New Jersey’s wage increase.
Econometrics · policy evaluation · replication
MSBA course preference MaxDiff
Impact: Compares counts, MLE, and Bayesian estimates to identify stable preference leaders.
MaxDiff · MNL · Bayesian estimation
A little more context
Finance taught me to ask where value comes from.
Analytics lets me test the answer.
I care about the bridge between a technically correct model and a useful decision. That means defining the action first, choosing a method that fits the evidence, and communicating the tradeoffs without hiding behind metrics.
Outside the notebook, I am refining coursework into durable portfolio work and learning how strong analytics teams turn uncertainty into momentum.
More about my background →Hiring for an analytics role, or curious about my work?
Let’s talk · kelun0925@gmail.com