Lead Informaticist, Medicaid Pharmacy Forecasting

About the role

The Lead Informaticist, Medicaid Pharmacy Forecasting owns the drug- and market-level utilization forecast for every Medicaid state Humana supports. This role is responsible for producing a forecast that is robust, defensible, and decision-ready, and for clearly explaining to executive stakeholders where variance is coming from when actuals diverge from expectations.

Responsibilities

• Own Medicaid Drug- and Market-Level Forecasting End-to-End
• Design, build, and maintain forecasts for drug-level utilization (script counts, days supply, cost, mix) and market-level utilization for each Medicaid state Humana supports.
• Produce forecasts at the cadence required by the business (e.g., monthly refresh, ad-hoc scenarios, launch projections, annual planning).
• Maintain a forecasting framework that is transparent, reproducible, and version-controlled—so results are traceable and defensible to executive and partner audiences.
• Account for the Realities of Medicaid
• Explicitly handle state-by-state differences in: Data availability, completeness, and lag; Member cohort composition, eligibility patterns, churn, and risk mix; Benefit design, formulary, PDL, and prior authorization policies; Provider, pharmacy network, and dispensing patterns; Regulatory and reimbursement environment (FFS vs. MCO, carve-in/carve-out, supplemental rebate dynamics).
• Build forecasting approaches that accommodate new-state launches—including ramp curves, cold-start handling, lookalike methods, and Bayesian shrinkage or hierarchical approaches when state-specific history is thin or absent.
• Build Methodologically Sound, Robust Forecasts
• Select and apply the right method for the problem—e.g., classical time series (ARIMA, ETS, state-space), hierarchical and panel models, regression-based decomposition, machine learning (gradient boosting, regularized regression), Bayesian hierarchical models, and ensembles—with clear justification for the chosen approach.
• Quantify and communicate uncertainty (intervals, scenarios, sensitivity) rather than presenting point estimates alone.
• Stress-test forecasts against historical analogs, holdout periods, and reasonable counterfactuals; document assumptions explicitly.
• Establish and monitor forecast accuracy metrics (e.g., MAPE, WAPE, bias, calibration) at appropriate levels of granularity, and continuously improve methodology based on observed performance.
• Variance Explanation & Executive Communication
• Diagnose and clearly explain the drivers of variance to executive stakeholders—decomposing variance into intuitive components such as: Membership / cohort change; Mix shift (drug, category, channel, state); Unit cost / rate change; Utilization rate change; Launches, LOEs, policy changes, and one-time events.
• Build standing variance and attribution analytics so leaders see what changed, why it changed, and what it means every cycle—not just what the number is.
• Translate technical results into concise executive narratives that anticipate the questions VPs and SVPs will ask.
• Partner with Stakeholders and Drive Decisions
• Partner with clinical strategy, pricing, network, finance, actuarial, Medicaid market leadership, and state-facing teams to ensure forecasts reflect the best available business intelligence and operational reality.
• Support new-state launch readiness by producing pre-launch forecasts, sensitivity ranges, and post-launch tracking against expectations.
• Translate forecast insights into clear options and recommended actions—e.g., where to intervene, where to escalate, where to adjust assumptions—so leaders can act, not just observe.
• AI-Accelerated Forecasting & Tooling
• Leverage AI agents, copilots, and modern coding tools to accelerate model development, feature engineering, code review, scenario testing, and explanatory analytics.
• Operate hands-on in Databricks using Python, PySpark, and/or SQL, with reproducible pipelines and clear documentation.
• Establish good engineering hygiene for the forecasting codebase: parameterization, configuration, testing, and reusable components that support extensibility as new states, drugs, and scenarios are added.
• Elevate the Practice
• Document methodology, assumptions, and known limitations clearly so the forecast is understandable and maintainable by others.
• Mentor more junior analysts on forecasting technique, variance decomposition, and executive communication.
• Stay current on changes in Medicaid policy, and pharmacy market dynamics, and translate developments into forecast improvements.

Qualifications

• Bachelor's degree (or equivalent experience) in a quantitative discipline (Statistics, Economics, Data Science, Operations Research, Mathematics, Actuarial Science, Health Services Research, or related); advanced degree preferred.
• 5+ years of progressive quantitative analytics experience, with 3+ years specifically in forecasting (utilization, demand, financial, or comparable).
• Demonstrated experience producing drug-, product-, or market-level forecasts in a healthcare, pharmacy, payer, PBM, or comparable setting.
• Strong hands-on proficiency in Python, PySpark, and/or SQL, with the ability to build and maintain reproducible forecasting pipelines.
• Working knowledge of forecasting methods across classical time series, regression-based, and machine learning approaches; ability to choose and defend the right method for the problem.
• Demonstrated experience explaining forecast variance to non-technical executives in clear, decomposable terms.
• Comfort with leveraging AI agents and coding tools to accelerate analysis and iteration.
• Strong written and verbal communication skills, with a track record of translating quantitative work into executive-ready narratives.

Preferred Qualifications

• Experience working in Databricks or comparable lakehouse environments.
• Direct experience with Medicaid pharmacy data and an understanding of state-by-state operational, regulatory, and data realities.
• Familiarity with handling cold-start / new-market launches (e.g., hierarchical models, lookalike approaches, Bayesian shrinkage).
• Experience with uncertainty quantification (prediction intervals, Bayesian methods, scenario modeling).
• Familiarity with pharmacy-specific dynamics: launches, LOEs, biosimilars, GLP-1 category disruption, formulary/PDL change impacts, and PA policy effects.
• Experience standing up standing variance/attribution analytics that explain "what changed and why" each cycle.
• Track record of partnering directly with finance, actuarial, clinical, and market leadership teams.