Applied Scientist, Forecasting

Zillow Zillow · Consumer · Bangalore, India

Applied Scientist role focused on designing, evaluating, and deploying statistical, econometric, and machine learning methods for forecasting problems within Zillow. The role involves the end-to-end modeling lifecycle, from data engineering and feature creation to model deployment, monitoring, and explainability, with a strong emphasis on translating forecasts into actionable business insights and recommendations for senior leadership. Collaboration with cross-functional teams and improvement of shared forecasting tools and data pipelines are also key responsibilities.

What you'd actually do

  1. Design and evaluate statistical, econometric, and machine learning methods for forecasting problems. Your work will play a central role in increasing forecast accuracy and accelerating delivery timelines
  2. Build backtesting and validation frameworks to assess forecast accuracy, stability, and downstream forecast impact.
  3. Own the end‑to‑end modeling lifecycle, including scoping, feature engineering, model development, experimentation, deployment, monitoring, and model explainability.
  4. Translate forecasts into clear insights and recommendations for senior leadership, helping stakeholders understand drivers, uncertainty, and trade‑offs that guide Zillow’s strategy.
  5. Quantify uncertainty and clearly communicate model confidence, limitations, and trade-offs to technical and non-technical audiences.

Skills

Required

  • Advanced degree (PhD or Masters) in Economics, Statistics, Operations Research, Data Science, Computer Science, Econometrics, Mathematics, or a related quantitative discipline with 3+ years of experiences in professional applied scientist roles
  • Strong experience with time-series forecasting, nowcasting, econometrics, or related quantitative modeling techniques.
  • Strength in data engineering principles to help envision efficient data solutions at scale.
  • Experience explaining complex models and analytical concepts to stakeholders with non-technical backgrounds using clear takeaways and practical business framing.
  • Experience building applied statistical or machine learning models that support real business decisions in production settings.
  • Proficient in Python and SQL for data analysis, model development, validation, and deployment.
  • Comfortable working with incomplete, delayed, or noisy real-world data and designing robust estimation strategies.

What the JD emphasized

  • machine learning methods for forecasting problems
  • end‑to‑end modeling lifecycle
  • model development
  • model explainability
  • time-series forecasting
  • business decisions in production settings

Other signals

  • forecasting
  • machine learning
  • statistical methods
  • econometrics
  • time-series
  • production settings