Senior Data Scientist

Responsibilities

• Analyze large-scale Komodo healthcare data using SQL and Python or R to assess projection model inputs, outputs, and behavior across therapeutic areas and geographies.
• Design and run evaluation studies — constructing benchmarks, consistency and uncertainty quantification — and build diagnostic tooling that makes model behavior fast to assess and debug.
• Triage and investigate questions from customers and customer-facing teams; reproduce reported issues, form hypotheses, and isolate root causes through systematic analysis.
• Prepare and deliver clear methodological explanations — briefs, walkthroughs, and direct customer discussions — that translate model mechanics into answers customers can act on.
• Review and contribute to model code, specifications, and design documents in partnership with product engineering; validate changes before and after release.
• Maintain methodology documentation and evaluation reports; track model behavior across data refreshes and releases, flagging emerging issues early.

Requirements

• Graduate training (MS/PhD) in Statistics, Data Science, or a related quantitative field, or equivalent applied experience.
• 5+ years of industry experience applying statistical modeling and estimation to large-scale real-world healthcare data.
• Deep familiarity with US healthcare claims data — the coding ecosystem (ICD, CPT/HCPCS, NDC), how drugs are billed across medical and pharmacy benefits, the distinction between open and closed data sources and its analytical implications, and how common data-generation and processing issues manifest in downstream analytics.
• Experience evaluating model quality when ground truth is limited, with the creativity to design rigorous validation approaches under that constraint.
• Strong diagnostic and critical-thinking skills: able to take an ambiguous question ("this number looks wrong"), form hypotheses, and systematically isolate root cause with minimal oversight.
• Comfort engaging directly with customers on technical topics — listening well, explaining methodology accessibly, and maintaining credibility under scrutiny.
• Working proficiency in SQL and in Python or R for large-scale data analysis and statistical modeling.

Nice-to-Haves

• Direct experience with projection or extrapolation methodology — estimating population-level volumes from partial samples (e.g., sample-to-national projection, market sizing, weighting and calibration) — is a strong plus.
• Experience with Bayesian statistics and its application to real-world estimation problems.
• Fluency with AI-assisted development tools (coding copilots, agentic workflows) to accelerate analysis, prototyping, and documentation.