Job DescriptionAt Public Storage, data science sits right next to the decisions. Prices move every day across thousands of locations, and our models help decide them. As a Data Scientist II, you'll be the go-to data scientist for one part of the business: customer and revenue, capital allocation (acquisitions, development and finance), or operations. You'll be part of their planning, shape the questions, and build the models they rely on.
We're a small, senior team that works like a startup inside an S&P 500 company. Ideas get to production quickly, you own your work, and it shows up in the business rather than in a slide deck.
What you'll do
- Build and own production models. Pricing and elasticity, promotions, marketing effectiveness, demand forecasting, retention, lifetime value and recommendation. You'll write production-grade Python and SQL, and use LLMs, embeddings and RAG where they're the right tool.
- Be the data science partner for your business area. Join their planning, turn ambiguous questions into well-scoped problems, and present results as decisions and dollars, not model metrics.
- Run experiments that change what we do. Design and analyze A/B tests, holdouts and quasi-experiments. Set the success metric before launch and report the honest result after it.
- Build once, reuse everywhere. Work from our shared customer and real estate data and certified metrics, and add reusable models to a shared catalog so the next problem starts faster.
- Grow the team around you. Mentor analysts and junior scientists through code review and pairing, and back up a teammate's business area when needed.
- Own the full model lifecycle. Versioning, retraining, drift monitoring and documentation, meeting our model governance standards before anything goes live.
What success looks like
- Business partners depend on your models every week.
- Your experiments change what the company does, not just what it knows.
- Junior scientists get better because of your code reviews.
Qualifications- Bachelor's degree in a STEM or quantitative field, such as statistics, computer science, economics, engineering or applied mathematics.
- 3+ years in a production data science role, or 2+ years with a Master's, or 1+ year with a PhD.
- Strong SQL and Python in production settings, not notebooks alone.
- Applied statistics in both frequentist and Bayesian frameworks, plus experimental design and causal inference.
- A track record of shipping models that drove measurable business outcomes.
- Experience with a cloud data platform. Google Cloud and BigQuery preferred; Snowflake, AWS or Azure considered.
- Able to turn an ambiguous business question into a scoped problem, and explain uncertainty and trade-offs to non-technical partners.
Preferred
- 5+ years of production data science in pricing, revenue management, marketing, real estate, retail or consumer analytics.
- The modern ML stack: scikit-learn, XGBoost/LightGBM, PyTorch, PyMC, Prophet, and experiment tracking with MLflow or Vertex AI.
- MLOps experience: model registries, CI/CD, scheduled retraining and monitoring.
- dbt and orchestration tools such as Airflow or Cloud Composer.
- Hands-on experience with LLM applications, vector search, RAG or agentic AI workflows.
Additional InformationWorkplace- One of our values pillars is to work as One Team and we believe that there is no replacement for in-person collaboration but understand the value of some flexibility. Public Storage teammates are expected to work in the office five days each week with the option to take up to three flexible remote days per month.
