Responsibilities
- Lead the design and implementation of advanced causal inference and statistical frameworks to measure and forecast the effectiveness of Pearl’s clinical products and operational services.
- Architect Causal Frameworks: Design and build the scalable systems required to conduct rigorous impact analyses, moving beyond simple correlations to isolate the true "Pearl Effect" on patient populations. Set the technical bar for how we handle complex data challenges, including non-randomized treatment assignment, selection bias, and compounding intervention effects.
- Forecast Quality & Performance: Develop predictive models to issue forecasts for clinical quality measures (including eCQMs in MSSP and claims-based measures in REACH and LEAD). This includes establishing the "status quo" baseline to accurately quantify Pearl's incremental impact.
- Collaborate on Patient Risk: Partner with other Staff Data Scientists to refine and validate patient risk models, ensuring that "rising acuity" signals are integrated effectively into our performance evaluation loops.
- Lead Technical Execution: Partner with Engineering and Analytics to build robust data pipelines and ML infrastructure that support automated, repeatable performance measurement.
- Translate Insights for Action: Collaborate with Product and Clinical Operations leaders to turn complex statistical findings into actionable narratives that influence product roadmaps and practice coaching.
- Automate Model Lifecycles with AI Agents: Architect and oversee AI-driven agents that autonomously manage the end-to-end lifecycle of our statistical models — leveraging automation for continuous training, deployment, performance monitoring, and proactive model refreshes.
Requirements
- Advanced Quantitative Expertise: A graduate degree (Masters or PhD) in a quantitative field such as Statistics, Economics, Biostatistics, or Epidemiology, with 8+ years of experience in results-driven quantitative analysis.
- Deep Causal & Statistical Literacy: Proven experience implementing causal inference methodologies (e.g., diff-in-diff, synthetic control, propensity score matching) in real-world, messy data environments.
- Predictive & Forecasting Proficiency: Experience building time-series forecasts or risk-adjustment models, with a strong understanding of how to define and measure a baseline vs. an intervention effect.
- Full-Stack Data Science Skills: Expert-level proficiency in Python and SQL, with the ability to write production-quality code and design scalable data architectures.
- Architectural Thinking: Experience building or significantly contributing to scalable data science systems and infrastructure within a modern cloud environment (AWS, Snowflake, dbt).
- Exceptional Communication: The ability to explain the nuances of a p-value, a risk score, or an identification strategy to a non-technical audience.
Nice to Have
- Healthcare Quality Expertise: Deep familiarity with eCQMs, HEDIS, or claims-based quality measures within MSSP, ACO REACH, or similar CMS programs.
- Thought Leadership: Experience publishing peer-reviewed research or presenting complex scientific findings at industry conferences.
