Model Behavior Engineer

Working remotely, the full-time Model Behavior Engineer will focus on understanding decision-making processes, defining and measuring reasoning quality, and improving AI system behaviors to enhance user decision-making in scientific R&D and commercial strategy. Key responsibilities Understand the decisions being supported by collaborating with users and domain experts to identify objectives and constraints Define and measure good reasoning by building evaluations and investigating the effectiveness of improvements Design and implement reliable software systems that facilitate reasoning workflows and evaluations Required qualifications Experience with rigorous thinking about ambiguous problems, potentially from research or quantitative fields Curiosity across various domains and the ability to communicate complex ideas clearly Strong experimental instincts and the ability to devise useful tests in the absence of established benchmarks Proficiency in coding, particularly in Python or a comparable language, with experience in working with data and model APIs Practical familiarity with language models and their capabilities, along with strong engineering judgment in system design