Overview
Lead Data Scientist on the Measurement Science team, providing decision science that gives leaders defensible, statistically grounded evidence on campaigns, product changes and strategic initiatives.
What you'll do
- Design, build and advise on appropriate measurement, experimental or quasi-causal designs to improve the effectiveness of business decisions and results.
- Provide leadership with clear insights into the strengths and limitations of measurement approaches.
- Identify and evaluate key performance metrics, including validation of proxies and leading indicators.
- Contribute to the design, review and engineering of repeatable measurement methodologies and systems.
- Identify languages, tools and methodologies that can bring efficiencies and rigor to the team.
- Apply best practices to develop statistical and machine learning/AI techniques that enable rigorous data-driven decisions.
- Mentor and upskill junior data scientists.
What you'll need
- A minimum of 3 years demonstrated experience applying causal inference methodologies (experimental design, propensity score matching/weighting, double ML, cluster randomization, doubly robust estimation).
- Comfortable consulting with and educating non-technical business partners on a variety of measurement methods.
- Strong foundation in statistical inference, including hypothesis testing, generalized linear modeling, variance reduction techniques (CUPED) and Bayesian inference.
- Experience in statistical and machine learning fundamentals, including model evaluation, feature engineering, and supervised/unsupervised learning principles.
- Proficiency in Python and/or R for data analysis, data manipulation, and model development.
Nice to have
- Experience with dashboards or visualization tools (Power BI, Tableau, etc.).
- Familiarity with software engineering best practices (version control, code review, testing frameworks).
- Experience using CI/CD pipelines or modern development workflows for deploying data products, experiments, or analytics automation.
- Experience with agentic coding tools and methodologies (e.g. GitHub Copilot, Claude Code, spec-kit).
- Experience evaluating LLM and agentic systems.
- Proficiency in distributed compute systems (e.g. PySpark).
Details
- Location: US - Remote.
- Compensation: $134,000.00 - $209,750.00 annually, based on experience and qualifications.
- Background investigation required for the successful candidate.
- Allstate generally does not sponsor individuals for employment-based visas for this position.
Read the full description and apply on the company’s own careers page.