Description
Haven Safety is hiring an AI/LLM Engineer to build and operate production reasoning systems for its enterprise safety-learning platform. The role owns multi-step LLM pipelines, agent orchestration, Neo4j and vector retrieval, evaluation, tracing, monitoring, failure handling, and model selection across OpenAI, Anthropic, and Google. It requires production LLM systems, agentic workflows, LLM evaluation, retrieval judgment, production ownership, model tradeoff judgment, and strong Python; Neo4j and Cypher experience is strongly preferred. The role reports to the CTO, is part of the founding team, and does not provide visa sponsorship.
