• Dual Service Strategy (Inference & Training): Define the roadmap for two core service pillars: • Training Services: Orchestrating elastic, cost-optimized compute (GPU/CPU) for model training and experiment tracking. • Inference Services: Managing the deployment of models into high-availability, low-latency API endpoints. • Ease of Development & Deployment: Radicalize the user experience for ML Engineers. You will build self-service "push-button" deployment workflows that abstract away the complexity of Kubernetes and cloud networking. • Lineage & Reproducibility: Ensure every model has a clear "paper trail." You will define how we capture the lineage between data versions, training code, and production artifacts—a critical requirement for Biopharma compliance. • Observability & Governance: Build the tools to monitor model health in production. This includes infrastructure-level metrics (latency/memory) and model-level observability (drift/performance) to ensure system reliability. • Technical Stakeholder Engagement: Partner with Scientific IT and Platform Engineering to ensure our services integrate seamlessly with existing enterprise identity (IAM) and security frameworks. • Backlog & Execution: Act as the "CEO of the Service," translating complex infrastructure needs into clear, actionable epics and user stories for a high-performing engineering team.