• Map DPD opportunities for AI: Identify and size use cases across drug product operations (execution insights, investigations support, batch record intelligence, documentation acceleration, knowledge retrieval, exception management, production performance). • Build and ship solutions: Deliver POCs/MVPs hands-on - integrating GenAI with operational data and controlled documentation in secure, controlled, and adoption-ready ways. • Roadmap and value cases: Own a DPD AI roadmap with value sizing, dependencies, and adoption plans; align priorities with DPD stakeholders. • Balance quick wins vs foundations: Build what is needed now while shaping and standardizing patterns that will scale with evolving enterprise data/AI infrastructure (reusable components, integration standards, evaluation/monitoring, agent orchestration patterns). • Vision and transition plan: Define the DPD AI north star and a staged plan from POCs to reusable capabilities to scaled solutions as infrastructure matures. • Quality-aware design: Build with regulated realities in mind (traceability expectations, controlled content boundaries, auditability of outputs where required). • Drive ChatGPT adoption: Create role-based prompt kits, playbooks, SOP-adjacent guidance, and enablement sessions; measure adoption and iterate. • Stay current and apply responsibly: Continuously monitor the rapidly advancing AI landscape and selectively bring in relevant capabilities that are safe, practical, and measurable in DPD operations.