Agentic AI Evaluation &; Quality Engineer

Role overview We're hiring an engineer to investigate and improve the behavior of a production agentic AI invoice-extraction application. The role focuses on understanding why individual documents produce unexpected classifications, routing decisions, mandatory fields, or extracted outputs-and turning those findings into durable fixes and regression tests. Must have Experience testing, evaluating, or debugging LLM, ML, document-processing, or other probabilistic systems. Understanding of prompt…