Senior AI Agent Engineer - Lead, Enterprise Workflows

EY; Las Vegas (NV); Hybrid; USD 144,000 - 359,000; Full time

Job summary

EY Challenger is building the firm’s next generation of AI-enabled services. This role focuses on long-running agent execution, enterprise integrations, and difficult production failures, balancing determinism and autonomy while maintaining safety and control.

You will lead the engineering of agent tools, contexts, and runtimes that perform reliably at scale, with collaboration across product, security, and evaluation teams. A hybrid onsite requirement and travel up to 50% apply.

Qualifications

• Bachelor’s degree in a relevant field.
• 9-12+ years of relevant experience.
• Hands-on experience shipping large-language-model apps or agents.
• Proficient in Python or TypeScript.

Responsibilities

• Design agent architectures with deterministic steps and appropriate complexity.
• Build dependable connectors and interfaces with clear contracts.
• Engineer agent behavior across planning, routing, delegation, and handoff.
• Create durable context combining instructions, state, memory and feedback.
• Bound agent authority with scoped identities and credentials.
• Evaluate complete trajectories with tests and verified outcomes.
• Operate long-running execution with durable state and safe retries.
• Make behavior observable with traces and replay for debugging.
• Protect user control with intervention points for safety.
• Collaborate across product, platform, security and evaluation teams.

Skills

Python

TypeScript

Prod engineering

Agent architecture

Applied AI

Evaluation discipline

Security operability

Product judgment

Distributed execution

Identity controls

Education

Bachelor’s degree

Tools

MCP standards

Knowledge graphs

Job description

EY Challenger is building the firm’s next generation of AI-enabled services. This role focuses on long-running agent execution, enterprise integrations, and difficult production failures, balancing determinism and autonomy while maintaining safety and control.

You will lead the engineering of agent tools, contexts, and runtimes that perform reliably at scale, with collaboration across product, security, and evaluation teams. A hybrid onsite requirement and travel up to 50% apply.