Projects · Learning shelf · LinkedIn
I am building toward Applied / Product Data Science with strong ML Engineering skills. I care about the full path from a product question to a trustworthy decision: define the metric, validate the data, model the uncertainty, test the intervention, and ship a reproducible system.
I am pursuing an M.S. in Spatial Economics and Data Analysis at the University of Southern California (expected 2027), bringing an econometrics lens to product experimentation and applied machine learning.
Each case study is rebuilt from my own analysis and excludes private course material, restricted data, and unverifiable claims. Projects are linked only after passing a reality, license, reproducibility, and documentation review.
- Conversion Intelligence — acquisition scoring with a prediction-time contract: AP 0.133 versus 3.2% prevalence and 5.25× lift at the top 5%. An illustrative cost-sensitive audit separates decision loss from model score, and the stronger final-session model is marked non-deployable because its features arrive too late.
- Lifecycle Email Experimentation — a messaging case study that separates a limited retrospective source audit from a public-safe synthetic implementation with fixed windows, multiplicity correction, a pre-treatment negative control, and guardrails. Only 1 of 24 source funding-snapshot differences survived correction (+0.311 pp; adjusted p ≈ 0.011); no cadence winner was supported.
- Review Sentiment Reliability — a clean-room, synthetic-only reliability study for a rating-derived text proxy. Contract 2.0 enforces leakage-aware group/time evaluation, repeated null controls, serving stress, and train/serve parity. Because rating defines the target, the model may be redundant when ratings are visible; it makes no real-data, causal, or production-value claim.
| Frame the decision | Separate prediction from causality | Build for review |
|---|---|---|
| Start with the user, metric, prediction time, and cost of error. | Use observational models for ranking; use experiments for intervention claims. | Add data contracts, tests, CI, model cards, and honest limitations. |
| Depth I am developing | Breadth I am building | Long-term direction |
|---|---|---|
| Experimentation · Causal inference · Applied ML | SQL · PySpark · MLOps · Cloud · LLM systems | Applied / Product Data Scientist → Applied Scientist / ML Scientist |
I treat this as a roadmap, not a wall of skill badges. A technology appears as a demonstrated strength only after a project makes the design choices, limitations, and evidence visible.
A SQL/PySpark point-in-time feature pipeline with data-quality and batch/serving parity tests. This is planned evidence, not a current skill claim.
Measure carefully · Build responsibly · Improve in public