Enterprise AI Agent Systems Engineer (On-Site)

Polluxa, Inc.; San Francisco (CA); On-site; USD 180,000 - 260,000; Full time

Job summary

Polluxa, Inc. in San Francisco is seeking a highly technical Enterprise Agent Systems Engineer to build and deploy advanced agent systems for large enterprise environments.

This role blends software engineering with applied AI research, requiring you to work with customer researchers and engineers to deliver production-grade data pipelines, post-training workflows, and robust evaluation infrastructure.

Qualifications

• Strong software engineering fundamentals and demonstrated ability to ship production systems.
• Hands-on experience building agent systems, AI applications, or technical forward-deployed solutions.
• Ability to personally write and own production code.
• Strong understanding of evaluation, agent behavior, and production reliability.
• Ability to work across technologies and learn new stacks as required by the problem.
• Strong communication skills with technical enterprise stakeholders.

Responsibilities

• Build and ship production agent systems for complex enterprise workflows.
• Develop data pipelines, orchestration systems, and evaluation harnesses.
• Run and analyze post-training workflows for customer-specific models and agents.
• Work directly with customer researchers and engineers.
• Evaluate whether performance issues originate in the model, environment, or evaluation framework.
• Translate customer workflows into robust, testable AI systems.
• Own technical deployments from implementation through production use.

Skills

Software engineering

Agent systems

Production code

Evaluation reliability

Cross-stack learning

Stakeholder comms

Job description

Polluxa, Inc. in San Francisco is seeking a highly technical Enterprise Agent Systems Engineer to build and deploy advanced agent systems for large enterprise environments.

This role blends software engineering with applied AI research, requiring you to work with customer researchers and engineers to deliver production-grade data pipelines, post-training workflows, and robust evaluation infrastructure.