AI Systems Engineer - Secure Execution - Senior

EY•Hartford, DC

3d•$106,900 - $200,600•Hybrid

About The Position

We are seeking AI Systems Engineers to own the security and trust fabric of EY’s AI-native platform, comprising the identity, secrets, cryptographic, and attestation layer that makes agentic AI workloads deployable in highly regulated environments. This role owns the enforcement mechanisms that allow EY to prove every workload is identity-bound, every secret is protected, every node is trusted, and every deployment is attestable and audit-ready. This is the strategic differentiator of the platform. The ability to deploy AI workloads in regulated industries and prove they are secure, economically bounded, and auditable depends on the trust fabric this role builds. It is the technical enforcement layer behind the AI Integrity / Security control authority and much of the platform’s governance capabilities, spanning consistently across cloud, on-prem, edge, and air-gapped environments.

Requirements

Deep expertise in workload identity, secrets management, PKI, and cryptographic lifecycle at production scale across multiple environments.
Strong understanding of confidential compute, trusted execution environments, hardware roots of trust, and remote attestation.
A security-first mindset: thinking in terms of provable trust, blast radius, least privilege, and cryptographic attribution rather than perimeter or convenience.
Ability to encode trust and compliance directly into infrastructure so that security is enforced by the platform, not by manual review.
Comfortable operating across cloud, on-prem, edge, and air-gapped environments with consistent identity and trust mechanisms.
Strong communicator able to explain trust, identity, and attestation tradeoffs to engineers, architects, auditors, and leadership.
Orientation toward auditability and evidence: able to translate regulatory expectations into technical controls and demonstrable proof.
Bachelor's or Master's degree in Computer Science, Security, or related technical field, or equivalent experience.
8+ years in security engineering, identity/PKI, or trust infrastructure, with hands-on production ownership.
Deep, hands-on expertise with workload identity (SPIRE/SPIFFE), IAM (Keycloak/Entra ID), and secrets management (OpenBao/Vault).
Strong grounding in PKI, X.509 certificate lifecycle (cert-manager), key management, and transit encryption.
Working experience with confidential compute and hardware attestation (TDX, SEV-SNP, SGX, NVIDIA CC, or equivalents) and secure boot (Intel TXT).
Experience delivering identity and secrets consistently across multi-tenant, multi-environment (cloud/on-prem/edge/air-gapped) platforms.
Proven track record operating under compliance, security, or regulatory constraints with audit-grade evidence requirements.
Ability to define clean ownership boundaries and consumption contracts with platform, data, and runtime teams.

Nice To Haves

Familiarity with secure DPU architectures (DOCA) and hardware root-of-trust / boot-chain designs.
Experience integrating identity and attestation into service mesh, policy engines (OPA), and API gateways.
Exposure to AI/ML workloads and the specific trust challenges of confidential AI inference (models/secrets inside enclaves).
Experience producing attestation and compliance evidence for external auditors or regulators.
Relevant certifications (e.g., CISSP, cloud security specialties) or demonstrable equivalent depth.
Exposure to regulated industries (financial services, tax, healthcare, risk).

Responsibilities

Own workload identity and secrets management: SPIRE/ODIS, Keycloak/Entra ID (IAM), OpenBao (secrets store), cert-manager (X.509 lifecycle), PKI issuers/roots, and transit encryption — propagated consistently across every environment and tenant.
Build confidential compute environments: TEE (TDX/SEV-SNP/SGX/TrustZone/CCA/NVIDIA CC), Intel TXT boot security, and secure DPU architecture (DOCA), so environments are isolated, attestable, and audit-ready.
Establish the platform-wide identity model so every workload, agent, and service carries a verifiable, propagated identity that flows through telemetry, cost attribution, and policy enforcement end-to-end.
Own the cryptographic lifecycle: issuance, rotation, revocation, and expiry of certificates, keys, and roots, with zero manual, untracked secrets and no long-lived credential sprawl across tenants.
Enforce attestation policy: which nodes, enclaves, and workloads are trusted, how trust is proven at boot and at runtime, and how attestation evidence is captured for audit.
Partner on a dotted-line basis with Enterprise Security / Cloud Platform / SRE to ensure independent review, alignment to enterprise trust standards, and audit readiness in regulated client contexts.