Technology Architect Assistant Director

EY•Atlanta, DC

7d•$131,900 - $280,200

About The Position

The Artificial Intelligence (AI) Identity Architect partners across technology and business teams to design and govern the integration of Agentic AI, automation, and intelligence into EY’s global Identity and Access Management (IAM) ecosystem. The role collaborates with Security Architects, Development Leads, Engineering, Operations, and internal business teams to ensure AI solutions are securely designed, compliant with IAM standards, and aligned with global technology and security strategies. The architect aligns business, operational, and security requirements and translates them into technical IAM capabilities. These include identity lifecycle management, authentication and authorization, session management, access requests and provisioning, access reauthorization, remote access, IAM APIs, role and entitlement models, and directory technologies. The role works with other architects and service owners to develop roadmaps that meet strategic and current requirements. The position also leads the design of Microsoft Entra Agent ID and Agent 365 capabilities, including agent identity models, governance, discovery and registration, access controls, delegation and consent, shadow-AI discovery, and remediation. It establishes architectural guardrails and assurance for AI and agentic systems, and partners with platform, data, Microsoft Purview, and Microsoft Defender teams to support governance, protection, auditability, and threat response.

Requirements

Bachelor’s or Master’s degree in Information Assurance, Computer Science, Information Systems, or a related field.
12 or more years of practical IT experience.
8 or more years of direct Information Security experience.
6 or more years of hands-on Identity and Access Management experience.
2 or more years of experience implementing or architecting AI/ML identity systems.
Proven experience providing architecture guidance and advisory services for varied environments and systems, including application and technology blueprints, roadmaps, optimization, and migration strategies.
Extensive experience implementing or architecting AI, machine learning, or automation technologies within IAM or cybersecurity domains.
Experience designing and integrating systems using Model Context Protocol (MCP) or equivalent AI orchestration standards.
Strong knowledge of Agentic AI architectures, including multi-agent systems, LLM orchestration, reinforcement-based agents, or decision engines.
Strong experience with enterprise IAM platforms such as Entra ID, SailPoint, CyberArk, and PlainID, and with identity analytics.
Ability to communicate and advocate security requirements and controls effectively for implementation by development teams.
Ability to communicate effectively, verbally and in writing, with all levels of management.

Nice To Haves

Experience designing or governing Agentic AI agents for workflow orchestration or decision automation.
Hands-on familiarity with Agentic AI frameworks such as LangChain, CrewAI, AutoGen, or similar.
Ability to explain complex AI and identity concepts to non-technical stakeholders.

Responsibilities

Partner with Architecture, Information Security, Client Technology, and Enterprise Technology teams to integrate AI-driven capabilities into enterprise IAM strategy in alignment with business goals, security policies, regulations, and technology roadmaps.
Architect AI-augmented solutions that improve the intelligence, automation, and adaptive capabilities of global IAM platforms.
Integrate AI features such as adaptive access control, predictive threat detection, and identity analytics into the IAM framework.
Define architectural standards, frameworks, and patterns for AI-assisted IAM services across on-premises, hybrid, and cloud environments.
Establish architectural guardrails and governance models for safe and auditable deployment of AI and agentic systems.
Define identity management strategies for AI agents, including their lifecycles, entitlements, and ownership relationships.
Advise senior leadership on applying Agentic AI to identity operations, user experience, and risk reduction.
Partner with data and platform teams to align with enterprise AI governance and model management frameworks.
Design the Microsoft Entra Agent ID agent identity model, including identity blueprints, parent/child inheritance, agent instances, sponsorship and ownership, and credentials for delegated versus application-only access.
Establish the Microsoft Agent 365 control plane across Registry, Access Control, Visualization, Interoperability, and Security, and define the operating model for its registry in the Microsoft 365 admin center.
Define a registration and discovery pathway that brings agents built with Copilot Studio, Microsoft Foundry, the Microsoft Agent Framework, the Agent 365 SDK, and third-party or open-source frameworks into a single authoritative registry.
Design Conditional Access for agents, risk-based responses using Microsoft Entra ID Protection, and network-layer controls for agent traffic.
Extend Microsoft Entra ID Governance to agent identities through access packages, access reviews, lifecycle workflows, sponsor notifications and transfers, and time-bound access with automated expiry.
Design delegation and consent for agents acting on behalf of users, including token exchange, scope minimization, and revocation paths.
Define shadow-AI discovery for locally run and cloud-hosted agents surfaced through Microsoft Defender, Microsoft Intune, and Agent 365, and establish remediation from discovery to governed identity.
Design systems and integrations using Model Context Protocol (MCP), agent-to-agent (A2A) patterns, and equivalent orchestration standards with authentication, authorization, and auditability built in.
Partner with Microsoft Purview and Microsoft Defender owners to ensure data protection, DLP, audit, and threat response cover agent activity end to end.
Define controls for internal policy, external regulation, and audit requirements for non-human identities, and evidence compliance through logging, reporting, and recertification.
Set standards for privileged agent access, including no standing privilege, short-lived credentials, just-in-time elevation, and least-privilege scope design.