Lead AI Architect – Universal Agent Control

EY; Jacksonville (FL); Hybrid; USD 126,000 - 230,000; Full time

Benefits offered by this job

Hybrid work model

Medical and dental coverage

Pension and 401(k)

Generous paid time off

Job summary

EY in the United States is seeking an experienced Lead AI Software Architect to lead a team of domain experts in designing and implementing AI Agent Control—a universal harness for managing agentic systems. You will drive identity/auth, post-training environments, and governance across AI capabilities.

The role requires extensive experience with multi-agent architectures, policy definition, and agent tooling, collaborating with Security, Data and Infra teams while influencing business

Qualifications

• Advanced degree in a technical field and 10+ years in software/AI engineering.
• Experience designing multi-agent systems and agent control architectures.
• Familiarity with AI native development tools and modern CI/CD pipelines.
• Experience with identity, auth, and secrets management in large systems.

Responsibilities

• Design control harness intelligence and multiagent systems.
• Architect agent identity/authentication for large-scale systems.
• Design post-training and tuning environments for agents.
• Develop synthesis models for AI open-source models.
• Implement and benchmark architectures against evals.
• Lead a team of experts in universal agentic control harnesses.
• Collaborate with AI, Security, Data and Infra teams for balanced outcomes.
• Support strategic business development and IP filings.

Skills

Advanced leadership

Analytical problem-solving

Cross-functional collaboration

Mentoring engineers

Emotional intelligence

Critical thinking

Stakeholder management

Education

Advanced degree in Electrical Engineering, Computer Science or Mathematics

Tools

AI code tools (Cursor/Claude Code/Qodo/Codium/Gemini)

NVIDIA NVAIE experience

AWS/Azure AI stacks

Job description

EY in the United States is seeking an experienced Lead AI Software Architect to lead a team of domain experts in designing and implementing AI Agent Control—a universal harness for managing agentic systems. You will drive identity/auth, post-training environments, and governance across AI capabilities.

The role requires extensive experience with multi-agent architectures, policy definition, and agent tooling, collaborating with Security, Data and Infra teams while influencing business