Product Data Scientist – Home Improvement

The role is a senior‑level product data scientist focused on home‑improvement lending. You’ll work hybrid in San Francisco, partnering with product managers and engineers to turn vague business questions into concrete, data‑driven actions.

What You'll Actually Be Doing

You’ll spend most of your day translating product hypotheses into structured analyses, designing and running A/B tests, and building the metrics framework that guides the product roadmap. Expect to write SQL and dbt models daily, keep event tracking reliable, and surface anomalies after launches. You’ll also prototype ML models (propensity, segmentation, pricing) and iterate on them with fairness and explainability in mind, all while communicating findings to senior stakeholders.

The Core Tech Stack

The non‑negotiables are solid SQL chops, Python (or R) for statistical work, and a hands‑on familiarity with dbt and Airflow‑style orchestration. You’ll need to be comfortable with hypothesis testing, power analysis, and the whole A/B testing lifecycle because the team lives by experimentation. Bonus points if you’ve already played with LLM‑assisted analytics or have built pipelines feeding tools like Amplitude or Segment.

Interview Expectations

1. Design an experiment to evaluate a new checkout flow for a home‑improvement loan product. The interviewer will probe how you set up control/treatment groups, calculate required sample size, handle sample‑ratio mismatch, and interpret both statistical and business significance. They’re looking for rigor, practical trade‑off thinking, and an ability to explain the experiment to non‑technical stakeholders.
2. Walk through building a propensity‑to‑purchase model that must satisfy fairness constraints. Expect to discuss feature engineering, model selection, validation metrics, and how you’d audit the model for bias and regulatory compliance. The hidden goal is to see if you can balance predictive power with explainability and ethical considerations.

Application Advice

Tailor your resume to echo the JD’s language: highlight “product analytics,” “A/B testing,” “hypothesis testing,” “power analysis,” “north‑star metric,” “cohort analysis,” “SQL,” “Python,” “dbt,” “Airflow,” “ML model validation,” “fairness,” and “LLM‑enabled workflows.” Mention any fintech or lending experience, and note that you can work Pacific Time and three days onsite in San Francisco. Embedding these keywords naturally will help you get past the ATS and signal that you’ve read the posting closely.