1x
AI Researcher
San Carlos, CA. Posted 5 months ago
The role
This person will advance NEO's intelligence by building AI systems, infrastructure, and data engines. They will work on model and data, data infrastructure, ML infrastructure, and evaluations to enable robots to learn from experience.
What you will do
- Advance robot capabilities through research, scaling data pipelines, optimizing training and inference throughput, or building evaluations
- Build infrastructure that multiplies team research velocity
- Ship research to production by owning the path from experimental result to deploy capability on robot hardware
- Contribute to a learning flywheel where robot experience leads to better models and more capable robots
Required qualifications
- Strong Python and PyTorch (or equivalent deep learning framework)
- Experience in large-scale codebases and data tooling and visualization
- Demonstrated experience in at least one area of the four pillars of AI: model and data, data infrastructure, ML infrastructure, or evaluation protocols
- Degree in Computer Science, Machine Learning, or a related field
- Track record of impact: published research, deployed production in modern AI systems, or infrastructure that measurably accelerated a team's work
Nice to have
- Graduate-level education or equivalent research experience
- Experience with distributed training frameworks (TorchTitan, DeepSpeed, FSDP/ZeRO)
- Experience with large-scale data processing pipeline and ETL systems spanning on-device, on-premise, and cloud infrastructure
- Experience with multi-modal generative models, world models, diffusion models, or autoregressive architectures
- Experience with inference optimization techniques: quantization (PTQ, QAT, INT8/FP8), CUDA/Triton kernel development, or serving systems (TensorRT or equivalent)
Benefits
- Health, dental, and vision insurance
- 401(k) with company match
- Paid time off and holidays
About the company
The 1X World Model Lab is an embodied AI research organization focused on pretraining foundation models to accelerate the emergence of embodied intelligence.