Rebuilding American manufacturing infrastructure through AI-native systems for the defense industrial base.
AI/LLM Integration Engineer
We are building the connective tissue between large language models and the physical world of production—inventory flows, supply chains, shop-floor scheduling, quality systems. This is not a chatbot wrapper. You will design agentic workflows that reason over structured ERP data, route decisions through retrieval-augmented pipelines, and surface operational intelligence to engineers and plant managers in real time. The work sits at the intersection of high-reliability software and frontier model capabilities, where latency, determinism, and auditability carry real weight. If you have ever wanted to see your inference graphs change how a factory floor moves, this is that role.
What they're looking for
- 3–10 years building production backend systems, with at least 2 years focused on LLM orchestration, prompt architecture, or agent frameworks (LangChain, LlamaIndex, or similar)
- Deep fluency in Python, FastAPI, and PostgreSQL; comfortable designing GraphQL APIs that serve both human operators and autonomous agents
- Experience grounding model outputs with vector databases and retrieval-augmented generation in domains where correctness matters more than creativity
- Strong opinions on evaluation—you know how to benchmark LLM pipelines beyond vibe checks and can build automated regression suites for agent behavior
- Familiarity with AWS infrastructure and an instinct for keeping inference costs predictable when scaling across thousands of concurrent reasoning calls