About this role
Advance NEO's intelligence by building AI systems, infrastructure, and data engines to enable robots to learn from experience, including multi-modal world models, large-scale data pipelines, distributed training/inference systems, and evaluation frameworks that connect pretraining metrics to real-world robot performance.
Skills for this role
PythonPyTorchDeep learning frameworksLarge-scale codebasesData toolingData visualizationDistributed training frameworksTorchTitanDeepSpeedFSDPZeROLarge-scale data processingETLOn-device infrastructureCloud infrastructureMulti-modal generative modelsWorld modelsDiffusion modelsAutoregressive architecturesQuantization (PTQ)Quantization (QAT)INT8FP8CUDATritonTensorRTGPU training and inferenceModel evaluation frameworksModel benchmarkingModel ranking systems''Tokenization''Robotics''Data annotation and curation'