Field AI Engineer: Build Agentic, Production-Ready Systems

Nexus Black; United States; On-site; USD 180,000 - 240,000; Full time

Job summary

Nexus Black, an AI-first business within IFS, seeks engineers to build agentic systems that reason, orchestrate, and produce traceable outputs for enterprise customers. You will ship production code from week one and bring learnings from field visits back into development.

You’ll design multi-step workflows, own evaluation, and ensure reliability with observability and error handling. A strong background in Python/TypeScript and model APIs is required, with a passion for practical deployments.

Qualifications

• Shipped systems using model APIs (Anthropic, OpenAI or similar).
• Built eval pipelines with golden datasets and regression suites.
• Strong software fundamentals: Python/TypeScript, Git, testing, CI.
• Able to reason about when deterministic code beats a model call.
• Curious about the physical world and user contexts such as field service environments.

Responsibilities

• Design and build agentic workflows with multi-step tool use and orchestration.
• Own evaluation: define correctness, build eval harnesses and golden datasets.
• Add production-grade features: observability, tracing, and robust error handling.
• Engage with customer rides and field visits to translate learnings into code improvements.
• Work with coding agents (Claude Code, Cursor) daily and continuously improve tooling.
• Contribute like an owner: PR reviews, design docs, and documentation.

Skills

LLM production experience

Python

TypeScript

Git

Testing

CI/CD

Deterministic code over model calls

Observability

Field/industrial curiosity

Tools

Claude

LangChain

LangGraph

LangSmith

Cursor

Job description

Nexus Black, an AI-first business within IFS, seeks engineers to build agentic systems that reason, orchestrate, and produce traceable outputs for enterprise customers. You will ship production code from week one and bring learnings from field visits back into development.

You’ll design multi-step workflows, own evaluation, and ensure reliability with observability and error handling. A strong background in Python/TypeScript and model APIs is required, with a passion for practical deployments.