Lead AI Architect & Enterprise Agent Control

EY; Nashville (TN); On-site; USD 144,000 - 329,000; Full time

Job summary

EY is hiring an experienced Lead AI Software Architect to lead a team of experts in designing and implementing AI Agent Control. You will own the development of universal agentic control harnesses, including tool usage, state management, environment control, policy application, and multiagent API design.

You will drive architecture choices, research new authorization methods, and mentor engineers in a fast-paced environment.

Qualifications

• Advanced degree required in a technical field (Electrical Eng, CS, Math).
• 10+ years of software engineering and AI experience.
• Experience with multi-agent system administration and modern secrets management.
• Familiarity with AI-native development tools and code assistants.
• Proficiency in scripting languages (Python, Bash, Typescript) and Go/Rust/Java.
• Understanding of AI-native design processes and identity/auth methods.
• Experience with modern IDEs, CI/CD, Git hosting (GitHub/GitLab/Azure).
• Experience with Nvidia NVAIE, AWS and Azure AI stacks.

Responsibilities

• Design control harness intelligence systems for multiagent setups.
• Architect agent identity/auth at scale in enterprise systems.
• Create post-training and tuning environments for AI agents.
• Develop synthesis models for open-source AI models.
• Implement and benchmark architectures against evals.
• Lead the team in building universal agentic control harnesses.
• Collaborate with AI, Security, Data, and Infra teams for outcomes.
• Support strategic business development for emerging tech.

Skills

Analytical skills

Leadership

Cross-functional collaboration

AI/ML domain knowledge

Communication

Education

Advanced degree in Electrical Engineering, Computer Science or Mathematics

Tools

Cursor

Claude Code

Qodo/Codium

Gemini

AWS Q

GitHub

Job description

EY is hiring an experienced Lead AI Software Architect to lead a team of experts in designing and implementing AI Agent Control. You will own the development of universal agentic control harnesses, including tool usage, state management, environment control, policy application, and multiagent API design.

You will drive architecture choices, research new authorization methods, and mentor engineers in a fast-paced environment.