Inference
Machine Learning Researcher
$250K–$350K
San Francisco · On-site · Full-time · Posted Jan 5, 2026
- Experience
- 3+ years of experience training AI models using PyTorch.
About the job
Conduct research and experiments in LLM post-training, training systems, and model architectures, then validate successful approaches and help deploy them in customer products.
What you'll do
- Research model architectures, learning methods, and techniques to improve model quality, capability, latency, and serving efficiency.
- Train models, develop distillation methods, and build benchmarks and evaluations to validate research findings.
- Collaborate with engineers to productionize successful research and document findings.
What they ask for
- 3+ years of experience training AI models using PyTorch.
- Deep understanding of transformer architectures, attention mechanisms, and model internals.
- Hands-on experience post-training LLMs using SFT, RLHF, DPO, or other alignment techniques.
- Experience with LLM-specific training frameworks and NVIDIA GPUs at scale; strong experimental methodology and ML fundamentals.
Nice to have
- Publications in ML venues.
- Experience with model distillation, LLM speed optimization, multimodal models, distributed training, or open-source ML projects.
Pay
$250,000 - $350,000 USD base salary
Tools
PyTorch, Hugging Face Transformers, DeepSpeed, Megatron, TRL, NVIDIA GPUs, SFT, RLHF
AI focus
LLM post-training research, including architecture and learning-method experiments, reinforcement learning for alignment and capability, distillation, and model evaluation.
Model-written summary · the original posting is authoritative