AI Platform Engineering Lead (Senior Manager)

EY is seeking a senior leader to build and operate the infrastructure and cloud-native platform underpinning its AI runtime. The role leads and develops the AI Platform Infrastructure Engineering team while hands-on delivering secure, reliable, and portable AI environments across cloud, on-premises, edge, client-managed, and air-gapped deployments.

Responsibilities:

• Lead and develop the AI Platform Infrastructure Engineering team: direct the day-to-day and longer-term work of AI Infrastructure and AI Systems Engineers, set priorities, allocate work, and remove blockers so the team delivers predictably; hire, mentor, and coach engineers, set engineering standards (design reviews, code reviews, runbooks), and grow technical depth and leadership capability across the team
• Build hands-on as a senior individual contributor: personally design, implement, and debug the hardest and highest-risk components of the platform — cluster bring-up, GPU and networking enablement, multi-tenancy, inference and routing — and lead by example on engineering quality
• Own the infrastructure delivery roadmap: sequence the build into phases that each put a usable capability into engineering teams’ hands, define and defend entry and exit gates, and report progress, risk, and trade-offs to senior leadership
• Own the infrastructure foundation and cloud-native platform: from Ubuntu/OS, BMC/bare-metal, DPU architecture, and NVAIE (GPU/Network/DCGM) through compute, Kubernetes and scheduling, cluster fabric and networking, multi-tenancy, and distributed compute — delivering foundational capabilities (Infrastructure Management, Kubernetes & Scheduling, Cluster Fabric Management) as a provisioned, patched, production-ready substrate for Agentic AI workflows and downstream runtime, data, and execution services
• Stand up and operate EY Agentic AI environments: across cloud (EKS/AKS/GKE), on-prem AI Factory (RKE2/NVAIE), edge (K3s), client-managed, and air-gapped deployment modes, maintaining one consistent stack contract across all targets and keeping the substrate modular and swappable so components can be replaced without rewriting consumers or creating vendor lock-in
• Own secure execution, inference, and routing: Ray Serve, vLLM/NIM/Triton, and NVIDIA Dynamo, with sandboxed execution (NVIDIA OpenShell) for isolated, safe model execution; and Envoy AI Gateway, semantic routing (vLLM-SR), model and prompt selection, and streaming response handling — directing each request to the right model under the right constraints
• Own operational excellence, resilience, and platform delivery: define SLOs, on-call and incident management practices, and lead post-incident reviews; own backup, disaster recovery, and cross-environment replication for high availability (Velero, CloudNativePG, Cilium ClusterMesh), patching, security hardening, and supply-chain hygiene; and partner with DevOps Engineers on CI/CD/CV (ArgoCD) and infrastructure-as-code / GitOps (Helm/OpenTofu) so environments are reproducible and drift-free
• Own GPU capacity, platform economics, and vendor selection: forecast demand, set quotas and cost attribution, and manage the infrastructure budget; evaluate commercial against open-source options, run proofs of concept, manage strategic vendor relationships, and work with procurement on long-lead hardware and licensing, making the commercial case for capacity to senior leadership