AI Agent Engineer

Outlier AI; Little Canada (MN); On-site; USD 90,000 - 130,000; Full time

Job summary

Outlier Ai in Minnesota is seeking a backend engineer to join our AI automation and systems integration team. You will work on production-grade services handling parsing, business logic, and reporting while helping train autonomous agents with human feedback.

You will collaborate across teams on multi-step workflows, live tool integrations, and security-aware system design, delivering concise technical observations to improve agent performance and reliability.

Qualifications

• 2+ years of backend engineering or AI automation experience.
• Experience building production-ready software with modular architecture.
• Proficiency in at least two major programming languages and SQL.
• Experience in live production environments with multi-turn interactions.
• Strong attention to detail and ability to provide concise feedback.
• Experience with multi-stage coordination workflows and live integrations is a plus.

Responsibilities

• We help train advanced generative AI systems by providing human feedback that improves autonomous agents.
• We support the development of LLMs that can act as proactive, multi-step agents.
• We contribute to projects that teach systems how to design, coordinate, and refine complex architectural workflows.
• We evaluate system behavior in realistic environments and identify subtle functional or security issues.
• We deliver clear, detailed technical observations that help improve agent performance and reliability.
• We collaborate with the team on orchestration-focused tasks and real-world integration challenges.

Skills

Backend engineering

AI automation

Systems integration

Python

JavaScript

Go

Java

SQL databases

Production systems

Attention to detail

Job description

Salary: $90,000 - 130,000 per year

Requirements:

• We bring at least 2 years of experience in backend engineering, AI automation, or integrating complex systems.
• We have demonstrated success building and supporting production-ready software with clear modular architecture, such as separate services for parsing, business logic, and reporting.
• We are proficient in at least two major programming languages, such as Python, JavaScript, Go, or Java, and we have worked with SQL databases.
• We have practical experience developing in live, real-world environments rather than mocked setups, including managing multi-turn system interactions.
• We demonstrate exceptional attention to detail and can provide concise, high-value technical feedback on sophisticated system behavior.
• We consider experience with multi-stage coordination workflows, live tool integrations, persistent state handling, session discovery, and identifying issues such as privacy leakage, authority escalation, or prompt injection to be strong advantages.

Responsibilities:

• We help train advanced generative AI systems by providing human feedback that improves autonomous agents.
• We support the development of LLMs that can act as proactive, multi-step agents.
• We contribute to projects that teach systems how to design, coordinate, and refine complex architectural workflows.
• We evaluate system behavior in realistic environments and identify subtle functional or security issues.
• We deliver clear, detailed technical observations that help improve agent performance and reliability.
• We collaborate with the team on orchestration-focused tasks and real-world integration challenges.

Technologies:

AIBackendSupportJavaJavaScriptPythonSQLSecurityMore:We are working on projects that help leading AI organizations improve their autonomous agents by training large language models with human feedback. Our mission is to shape the future of generative systems that can plan, coordinate, and optimize complex workflows in real-world settings. We welcome experienced software builders and orchestration-minded professionals who want to contribute to cutting-edge AI development. The role focuses on advanced agent training and system evaluation, with opportunities to work on sophisticated multi-step interactions and live integrations.last updated 43 week of 2026