Remote AI Process Engineer — Agent Orchestration & HITL

GE Aerospace; San Ramon (CA); On-site; USD 152,000 - 202,000; Full time

Benefits offered by this job

Medical, dental, vision

401(k) with match

Paid time off

Job summary

GE Aerospace is seeking a leader to identify, deconstruct, and re-implement high-value manual processes using AI agents, LLMs, and modern automation. You will own the full arc from discovery to production agent orchestration with appropriate human oversight.

You will break down expert workflows, replace legacy tooling, and design robust, audit-friendly HITL architectures that ensure safe, scalable automation across aerospace processes.

Qualifications

• Bachelor’s degree from an accredited university or college.
• Minimum of 5 years of experience in software/AI engineering with at least 2 years building LLM-based or agent-based systems.
• Demonstrated ability to extract tacit knowledge from domain experts (process mining, cognitive task analysis, or equivalent).
• Experience designing and deploying AI agent orchestration (multi-step, tool-using, with evaluation).
• Strong systems thinking — can model a process end-to-end before writing a line of code.
• Track record of replacing legacy systems without burning the house down.

Responsibilities

• Process Discovery & Knowledge Extraction.
• Embed with domain experts to observe and document actual workflows (not the documented ones).
• Distinguish genuine expertise and edge-case reasoning from ritual, habit, and cargo-cutting.
• Identify where "art" is pattern recognition that can be modeled, versus true judgment calls requiring human decision-making.
• Produce formal process models from informal, oral-tradition knowledge.
• Architect multi-step AI agent workflows that decompose complex expert tasks into verifiable stages.
• Define tool-use patterns, context management, and failure/fallback strategies for agents.
• Build evaluation frameworks that compare agent output against expert baselines.
• Implement confidence-gated escalation — the system knows what it doesn't know.
• Design the HITL topology: which decisions require human approval, review, or override.
• Define escalation thresholds, audit trails, and feedback loops that improve the system over time.
• Ensure experts transition from "doers" to "reviewers and teachers" without loss of engagement or institutional knowledge.
• Build calibration mechanisms so human reviewers stay sharp (not rubber-stamping).

Skills

Process discovery

Knowledge extraction

LLM-based systems

Agent orchestration

Human-in-the-loop

Systems thinking

Decomposition

Education

Bachelor’s degree

Tools

LLM pipelines

Vision models

Structured extraction

Job description

GE Aerospace is seeking a leader to identify, deconstruct, and re-implement high-value manual processes using AI agents, LLMs, and modern automation. You will own the full arc from discovery to production agent orchestration with appropriate human oversight.

You will break down expert workflows, replace legacy tooling, and design robust, audit-friendly HITL architectures that ensure safe, scalable automation across aerospace processes.