低精度量化/模型压缩算法工程师
上海
全职
职位描述
1、负责具身智能模型、机器人多模态感知/决策模型的轻量化压缩与低精度量化优化，落地端侧/边缘部署，在保精度前提下降低推理时延、显存占用与设备功耗。 2、主导量化体系迭代，包含FP8/INT8/FP4/INT4混合精度量化、PTQ/QAT量化训练，解决具身模型动态输入、时序特征、控制任务带来的量化误差、漂移、精度掉点问题。 3、开展模型蒸馏、结构化/非结构化剪枝、稀疏化、算子融合、推理图优化等压缩方案，针对机器人实时交互、导航、操控场景做专项性能迭代。 4、对接推理框架与硬件平台，完成模型转换、算子适配、内存与调度优化，打通训练压缩—量化校准—端侧部署全链路，输出标准化压缩量化工具与基线。
职位要求
1、硕士及以上学历，计算机、人工智能、自动化、电子信息等相关专业，3~5年模型量化、模型压缩、端侧AI优化经验，有具身智能/机器人模型优化经验优先。 2、精通各类低精度量化算法，熟练掌握PTQ、QAT、混合精度量化方案，熟悉剪枝、蒸馏、稀疏化等主流模型压缩方法，具备精度修复与误差调优能力。 3、熟练使用PyTorch/TensorRT/ONNX，熟悉NPU/RKNN等端侧推理部署流程，具备算子优化、推理加速、显存功耗优化实战经验。 4、熟悉多模态、时序感知、机器人决策控制类模型特性，能独立完成模型压缩量化方案设计、实验验证、指标复盘与工程落地。
投递
AGIBOT is an Embodied AI foundation model company developing both the intelligence layer and the corresponding robotic embodiments needed to bring general intelligence into the physical world. Founded in 2023, AGIBOT is building a full-stack Embodied AI platform across foundation models, data infrastructure, simulation, robotic hardware, and deployment. AGIBOT’s “Three Intelligences in One” architecture integrates Locomotion Intelligence, Interaction Intelligence, and Manipulation Intelligence into a unified embodied system. AGIBOT’s portfolio spans humanoid robots, quadrupeds, dexterous systems, and commercial cleaning solutions. In March 2026, AGIBOT announced that its 10,000th robot had rolled off the production line.
