Senior Director, US Decision Science & AI Innovation

Position Summary

This role owns the analytic and AI methods that turn data into commercial decisions across US Commercial — marketing mix and investment optimization, next-best-action and dynamic targeting, patient predictive triggering, and causal measurement — and their deployment into the business through enterprise capabilities. The role sets what constitutes a defensible method, decides what is certified enterprise method versus brand configuration, and is accountable for the scientific quality, reproducibility and explainability of results used in brand and portfolio decisions. This is a player-coach role, not an analytics oversight role — the post holder sets standards and develops the organization while working directly in the analytics, setting the technical bar by example.

Key Responsibilities

• Own the certified method portfolio — marketing mix and investment optimization, next-best-action and dynamic targeting, patient predictive triggering, and causal measurement — driving reuse across brands and eliminating duplicate bespoke builds.
• Own deployment of those methods through enterprise capabilities — decision engines and simulators, embedded field and omnichannel triggers, planning and measurement tools — and be accountable for adoption and sustained use.
• Operate as a player-coach: personally design, review and build, setting the technical bar through direct involvement rather than oversight alone.
• Set and enforce method standards — validation, versioning, documentation, reproducibility and explainability — and define what is certified enterprise method versus brand configuration.
• Lead the applied AI agenda across agentic AI, generative AI and synthetic data — including multi-agent orchestration, retrieval-augmented generation and evaluation pipelines — with a prototyping pipeline whose explicit criteria either graduate techniques into production or retire them.
• Ensure methods, capabilities and AI applications meet GSK Responsible AI, privacy and data-use requirements, including model risk assessment, human oversight of automated recommendations, and permitted-use compliance for patient-level data, with a model registry and ongoing monitoring in place before anything scales.
• Partner with data engineering and the Digital & Technology organization for the platform and engineering these capabilities run on, and with data management for certified inputs.
• Own build-versus-buy decisions, biased toward building quickly and efficiently in-house; buy where it accelerates delivery and the capability can be absorbed into the GSK ecosystem. Hold vendors to internal standards.
• Act as method authority for brand teams and Insights & Analytics, translating demand into a prioritized roadmap sequenced against published capacity and communicating uncertainty and limitations to executive audiences.

Basic Qualifications

• Bachelor’s degree; Masters and PhD preferred.
• 12+ years of experience with hands on modeling.
• Experience deploying predictive analytics machine learning algorithms, deep learning algorithms and/or information retrieval at scale.
• End to end model exploration, development, deployment and management.
• Experience developing insights and storylines through translation of analysis and modeling.
• Depth in commercial pharmaceutical methods — marketing mix modeling, next-best-action, patient-level predictive modeling or causal measurement — with accountability for results used in business decisions.
• Production generative or agentic AI experience in commercial pharma, under enterprise governance — shipped and in use.

Preferred Qualifications

• Build and deploy tools and methodologies to create solutions enabling analytics and measurement of sales and marketing activities (e.g., market-mix modeling, cross-channel attribution, omni-channel segmentation and customer targeting, patient pathways, omni-channel customer journey analysis and optimization etc.)
• End to end experience in retrieving data and developing an executive-level analysis through advanced analytics techniques.
• Strong written and verbal communication skills.
• Ability to work effectively across boundaries – both cultural and functional.
• Leadership capability is critical; you’ll work in a matrix, inspiring change rather than commanding action.
• Experience setting agentic AI vision and governance across an organization rather than a single team.
• Working command of current GenAI tooling — LLM APIs, prompting and context engineering, retrieval-augmented generation (vector stores, chunking, retrieval evaluation) and LLMOps evaluation pipelines.
• Familiarity with multimodal data — structured claims alongside unstructured clinical notes and lab signals — and with US commercial data assets such as IQVIA, Komodo, Symphony and Veeva.