Generalist
Software Engineer: ML Optimization
$260K–$350K
San Francisco Bay Area (San Mateo) or Boston (Somerville) · On-site · Full-time · Posted Feb 12, 2026
About the job
Optimize the training and on-robot inference stacks for embodied ML models, pursuing major performance gains across hardware, software, and ML.
What you'll do
- Optimize ML training and inference stacks.
- Implement ML, hardware, and software changes to achieve step-function performance gains.
- Identify and address bottlenecks across GPUs, kernels, networks, CPUs, and software.
What they ask for
- Proficiency in current ML techniques for training and inference optimization.
- Experience or capability optimizing transformer- and diffusion-based architectures.
Pay
$260K – $350K USD base salary
Tools
CUDA, torch, numpy, Python, NVLink
AI focus
Optimize training and on-robot inference for ML models, including transformer and diffusion architectures.
Model-written summary · the original posting is authoritative