Principal AI Architect - Director

Citigroup; New York (NY); On-site; USD 170,000 - 300,000; Full time

Benefits offered by this job

Medical, dental & vision coverage

401(k)

Life, accident & disability insurance

Wellness programs

Paid time off

Job summary

Citigroup is seeking a Principal AI Architect in New York to define and govern the enterprise architecture for autonomous agentic systems and multi-agent frameworks. This executive role co-owns enterprise blueprints, review gates, and playbooks, aligning field deployments with Arc Platform, R2D2 Gateway, Agentflow, and Helix across Citi’s technology footprint.

With 10+ years in software architecture and 3+ years in production LLM-powered apps, you’ll lead governance, establish standards for MCP

Qualifications

• 10+ years of progressive software architecture and engineering experience in enterprise, distributed systems, or financial technology environments.

Responsibilities

• Serves as the chief technical authority for autonomous systems, multi-agent frameworks, and hierarchical sub-agent architectures across Citi's global technology footprint.
• Authors and maintains enterprise reference architectures and blueprints for single-agent, hierarchical sub-agent, and Human-in-the-Loop (HITL) workflows, ensuring strict adherence across all delivery pods.
• Acts as the primary technical bridge in the Product/Platform Feedback Loop with the Arc Core, R2D2 Gateway, and Agentflow platform teams—translating field friction from LOB deployments into core platform features.
• Chairs the Agent Architecture Review Board (AARB) to conduct mandatory technical, architectural, and security reviews for all high-risk AI agents prior to UAT and Production promotion.
• Defines enterprise standards for Model Context Protocol (MCP) server integration, zero-trust boundary enforcement via Tiramisu proxies, and COIN M2M identity token delegation.
• Establishes horizontal governance, code quality, testability, and automated evaluation standards across all Forward Deployed Engineering (FDE) pods.
• Advises senior business and technology leadership on emerging AI paradigms, non-deterministic system validation, and foundation model capabilities.
• Appropriately assess risk when business decisions are made, demonstrating particular consideration for the firm's reputation and safeguarding Citigroup, its clients and assets, by driving compliance with applicable laws, rules and regulations, adhering to Policy, applying sound ethical judgment regarding personal behavior, conduct and business practices, and escalating, managing and reporting control issues with transparency, as well as effectively supervise the activity of others and create accountability with those who fail to maintain these standards.

Skills

Python

LLM-powered applications

Multi-agent systems

Distributed microservices

Executive communication

Golang

Security & compliance

Education

Bachelor's/University degree or equivalent

Master's degree in Computer Science/AI/ML

Tools

OpenShift

Kubernetes

OAuth2/JWT

API gateway management

OpenTelemetry

Splunk

Job description

Principal AI Architect

The Principal AI Architect is a senior executive technical authority position responsible for defining, establishing, and governing the enterprise architectural vision for autonomous agentic systems and multi-agent frameworks across Citigroup. The overall objective of this role is to serve as Citi's chief technical authority for autonomous systems and multi-agent frameworks, co-owning enterprise blueprints, review gates, playbooks, and the Product/Platform Loop bridging field implementations with core internal platforms (Arc Platform, R2D2 Gateway, Agentflow, and Helix).

Responsibilities

• Serves as the chief technical authority for autonomous systems, multi-agent frameworks, and hierarchical sub-agent architectures across Citi's global technology footprint.
• Authors and maintains enterprise reference architectures and blueprints for single-agent, hierarchical sub-agent, and Human-in-the-Loop (HITL) workflows, ensuring strict adherence across all delivery pods.
• Acts as the primary technical bridge in the Product/Platform Feedback Loop with the Arc Core, R2D2 Gateway, and Agentflow platform teams-translating field friction from LOB deployments into core platform features.
• Chairs the Agent Architecture Review Board (AARB) to conduct mandatory technical, architectural, and security reviews for all high-risk AI agents prior to UAT and Production promotion.
• Defines enterprise standards for Model Context Protocol (MCP) server integration, zero-trust boundary enforcement via Tiramisu proxies, and COIN M2M identity token delegation.
• Establishes horizontal governance, code quality, testability, and automated evaluation standards across all Forward Deployed Engineering (FDE) pods.
• Advises senior business and technology leadership on emerging AI paradigms, non-deterministic system validation, and foundation model capabilities.
• Appropriately assess risk when business decisions are made, demonstrating particular consideration for the firm's reputation and safeguarding Citigroup, its clients and assets, by driving compliance with applicable laws, rules and regulations, adhering to Policy, applying sound ethical judgment regarding personal behavior, conduct and business practices, and escalating, managing and reporting control issues with transparency, as well as effectively supervise the activity of others and create accountability with those who fail to maintain these standards.

Qualifications

• 10+ years of progressive software architecture and engineering experience in enterprise, distributed systems, or financial technology environments.
• 3+ years of hands‑on experience architecting and deploying production LLM-powered applications, multi-agent systems, and autonomous agentic workflows (e.g., LangGraph, AutoGen, CrewAI, Semantic Kernel, Agent Development Kit / ADK).
• Mastery of Python and/or Golang, OpenShift/Kubernetes, distributed microservices, OAuth2/JWT token delegation, API gateway management, and enterprise telemetry (OpenTelemetry, Splunk).
• Deep understanding of financial services compliance, Model Risk Management (MRM) standards, and deterministic validation methodologies for non-deterministic AI outputs.
• Demonstrated leadership and track record of chairing technical governance bodies and influencing enterprise-level architectural direction.
• Consistently demonstrates clear, concise, and persuasive executive-level written and verbal communication.

Education

• Bachelor's/University degree or equivalent experience
• Master's degree in Computer Science, AI/ML, or related technical field strongly preferred.

Job Family Group: Technology

Job Family: Architecture

Time Type: Full time

Primary Location: New York New York United States

Primary Location Full Time Salary Range: $170,000.00 - $300,000.00

In addition to salary, Citi's offerings may also include, for eligible employees, discretionary and formulaic incentive and retention awards. Citi offers competitive employee benefits, including: medical, dental & vision coverage; 401(k); life, accident, and disability insurance; and wellness programs. Citi offers paid time off packages, including planned time off (vacation), unplanned time off (sick leave), and paid holidays. For additional information regarding Citi employee benefits, please visit citibenefits.com. Available offerings may vary by jurisdiction, job level, and date of hire.

Anticipated Posting Close Date: Nov 30, 2026