Data Scientist — Denver, CO
Зарплата не указана
ОткликнутьсяAbout The Role
The role focuses on turning large-scale behavioral and transactional data into models that drive core product and business decisions. Day-to-day work spans causal inference, forecasting, experimentation design, and production modeling — not just dashboards.
You will partner directly with ML engineers and product teams to move analyses into production, owning projects from hypothesis framing through deployed impact. Findings from this role ship as features, pricing decisions, and model improvements, not slide decks.
Key Responsibilities
• Design and analyze A/B tests and quasi-experiments, including power analysis, sequential testing, and causal inference methods (difference-in-differences, synthetic control, uplift modeling)
• Build and validate statistical and ML models — gradient boosting, logistic regression, survival models, time-series forecasting — for churn, LTV, and demand use cases
• Develop reproducible analysis pipelines in Python (pandas, scikit-learn, statsmodels) and SQL, version-controlled in Git with peer review
• Collaborate with ML engineers to transition prototypes into production services, defining evaluation metrics and success criteria upfront
• Query and model data at scale using SQL against Snowflake or BigQuery, partnering with data engineering on warehouse design and data quality
• Communicate results to product and executive stakeholders through clear written analyses, dashboards in Looker or Tableau, and technical walkthroughs
• Contribute to shared tooling: experimentation frameworks, metric definitions, and reusable modeling libraries
What We Are Looking For
• 3–6 years of experience in data science, applied research, or analytics at a product-driven technology company
• Advanced SQL skills and strong Python proficiency, including scikit-learn, pandas, and at least one statistical modeling library (statsmodels, PyMC, or equivalent)
• Hands-on experience designing and analyzing controlled experiments at scale, with a solid grasp of statistical pitfalls (Simpson's paradox, novelty effects, interference)
• Track record of shipping at least one model or analysis into production that measurably changed a business decision or product behavior
• MS or PhD in statistics, economics, computer science, or a related quantitative field (or equivalent practical experience)
• Strong written communication — able to explain a complex causal analysis to a non-technical executive in one page
• Bonus: experience with uplift modeling or heterogeneous treatment effect estimation; cloud ML tooling (SageMaker, Vertex gX); CausalML/DoWhy; publishing or open-source contributions