具身数据评测产品经理 - Seed Model
Core
负责具身智能模型（机器人/VLA）的产品能力评估、评测标准建设及数据闭环，为模型迭代和规模化部署提供可信依据。
Role type
Product Manager for Embodied AI Model Evaluation
Builds
Evaluation frameworks, benchmarks, rubrics, and data pipelines for embodied models
Domain
Robotics / Embodied AI / Generative AI
Deliverable
production ML models
Required skills
Product thinking, system abstraction, model training/evaluation understanding, data analysis, experimental judgment, project planning, cross-functional coordination
Preferred skills
Experience with VLA, multimodal models, robot operations, imitation learning, reinforcement learning
Technologies
VLA, Multimodal models, Robotics, Imitation learning, Reinforcement learning
Responsibilities
Establish model capability requirements and evaluation standards for entry; Iterate evaluation systems and benchmark definitions; Analyze evaluation results to ensure reliability and support decision-making; Identify capability gaps and prioritize improvements; Drive data loops to convert evaluation findings into data and model iteration plans.
