Description
Software Mind is hiring an AI Agent Reliability Engineer to own the trust layer of a production Python-based AI copilot for oil-refinery and chemical-plant process engineers. The role focuses on agent reliability rather than LLM endpoint features, including assumption and grounding audits, chart-quality reviews, model-evaluation harnesses, latency and concurrency fixes, and regression testing. The engineer will investigate customer-reported misbehavior, diagnose root causes, implement fixes, and improve code quality in a large shared codebase, with work distributed across US and Australia time zones.
