Senior AI Engineer

Oakwell Hampton Group; San Francisco (CA); On-site; USD 160,000 - 260,000; Full time

Job summary

Oakwell Hampton Group is seeking a Senior AI Agent Engineer to architect and scale production-grade agentic systems for industrial data and processes. You will design autonomous agents that can reason over engineering data, call tools, manage project state, and execute multi-step workflows across quoting, procurement, scheduling, and shop-floor execution.

We value hands-on production experience, strong Python, and expertise in agentic design, memory architectures, and knowledge retrieval.

Qualifications

• Hands-on experience shipping modern AI/LLM systems to production.
• Strong Python and production backend skills.
• Expertise in agentic design patterns and constrained generation.

Responsibilities

• Design and deploy production-grade AI agents for industrial workflows.
• Engineer robust tool-calling, routing, planning, memory systems.
• Build extraction and multimodal pipelines for drawings, BOMs, PDFs.
• Implement evaluation, verification, and human-in-the-loop guardrails.
• Collaborate with operations and fabrication teams to fix real-world frictions.

Skills

Python

LLM systems

Agentic design

Memory management

Vector databases

Knowledge retrieval

Production backend

Tools

LangGraph

Job description

We are partnered with a fast-growing industrial technology company building the foundational operating system and agentic infrastructure for American manufacturing. They are combining software, modern AI, and a proprietary network of fabrication facilities to help aerospace, energy, and infrastructure leaders build complex hardware faster and more reliably.

They are looking for a Senior AI Agent Engineer to architect and scale the production-grade agentic systems sitting at the core of their platform. You will move beyond simple RAG and chatbots, building autonomous and semi-autonomous agents capable of reasoning over engineering and manufacturing data, calling tools, maintaining complex project state, and executing multi-step workflows across quoting, project management, procurement, scheduling, and shop-floor execution.

What You'll Do

• Design, build, and deploy production-grade AI agents for high-consequence industrial workflows.
• Engineer robust tool-calling systems, routing, planning frameworks, memory architectures, and multi-agent coordination (e.g., LangGraph or similar).
• Build sophisticated extraction and multimodal pipelines for unstructured engineering drawings, BOMs, specs, PDFs, and communications.
• Implement rigorous evaluation frameworks, verification loops, and human-in-the-loop guardrails to ensure output reliability and trace-back to source documentation.
• Partner directly with internal operations and fabrication teams to observe real-world friction points and rapidly turn them into reliable software solutions.

What We're Looking For

• Exceptional software engineering background with deep, hands-on experience designing and shipping modern AI/LLM systems to production.
• Strong command of Python and production backend infrastructure.
• Demonstrated expertise in agentic design patterns (tool use, planning, memory, state management) and constrained generation.
• Solid grasp of RAG, vector databases, semantic search, and knowledge retrieval architectures.
• Pragmatic engineering mindset: you care deeply about whether an AI system actually works reliably in the real world rather than just looking good in a demo.
• No manufacturing background required—just a genuine curiosity about how physical hardware gets built.

Why This Opportunity?

• High Ownership & Impact: Build software that directly coordinates and controls physical production processes in critical sectors like aerospace and energy.
• Cutting-Edge AI Challenges: Tackle complex state management, long-term memory, multi-step reasoning, and verification systems where failure isn't an option.
• Fast-Paced Environment: Work closely with founding leadership and operators with minimal bureaucracy and rapid feedback loops.