Data Scientist (Consultant)

Overview

In this role you will advance the Decision Intelligence capability in Finance Reinvention by building ML-driven forecasts from finance data. You will own forecasting models end-to-end, from framing and data prep to production integration and monitoring. You’ll collaborate with planning, data, and AI teams to deliver explainable, CFO-facing insights that improve planning accuracy and business outcomes. This is a hands-on, impact-driven opportunity to shape enterprise forecasting and decision support at scale.

Responsibilities

• Build forecasting models on client financial and operational data using classical, econometric, ML, or DL methods.
• Prepare and validate multi-source data, engineer drivers, and model hierarchies across products, entities, geographies or cost centers.
• Design back-testing and time-series cross-validation; evaluate accuracy, bias, stability, and business impact; reconcile forecasts across hierarchies.
• Run scenario and sensitivity analyses for CFO-level interrogation, including stress tests and counterfactuals.
• Produce variance explanations and commentary suitable for FP&A, including plan-versus-actual and driver attribution.
• Integrate models into the client planning cycle and EPM platform; collaborate on pipelines, APIs, model registry, deployment, drift detection, and retraining.
• Collaborate with AI Engineers on agentic workflows, variance alerting, and narrative generation while maintaining finance review.
• Document methods, data, assumptions, limitations, and validation evidence; measure impact on decision quality and forecast performance.

Key requirements

• Deep expertise in time series and forecasting methods, with seasonality and external regressors handling and knowledge of model trade-offs.
• Python and associated analytics stack; production or near-production deployment experience; SQL, Git, testing, and reproducible ML pipelines.
• Strong forecast evaluation skills including time-series cross-validation, back-testing, benchmarks, and uncertainty metrics.
• Experience with large, multi-source datasets and implementing data-quality checks in forecasting pipelines.
• Ability to explain model behavior to a finance audience and defend assumptions, uncertainties, and limitations.
• Ability to align models with planning calendars, adoption workflows, and measurable outcomes with FP&A and other stakeholders.
• Minimum 4 years of relevant professional experience.

• Clear communication with finance stakeholders
• Cross-functional collaboration
• Problem framing and analytical thinking
• Time series forecasting (classical and modern)
• Python and analytical stack; SQL; Git; reproducible pipelines
• Forecast evaluation, cross-validation, and bias/unpredictability assessment

…