Amazon
Principal Engineer: Annapurna Labs Training and Inference , Annapurna Labs, AWS Neuron SDK
$200K–$311K
Seattle, WA +1 · Principal · Full-time · Posted Oct 14, 2026
- Location
- Seattle, WA
Cupertino, CA - Experience
- 10+ years of non-internship professional software development experience and 7+ years with distributed training or RL systems at scale.
About the job
Leads hands-on engineering to optimize AI model training, post-training, and inference performance on AWS Trainium through the Neuron SDK, while setting technical direction across teams.
What you'll do
- Optimize model training kernels, distributed training collectives, post-training workflows, and inference serving for performance, cost, and latency.
- Make technical architecture and implementation decisions, including influencing hardware-software co-design and enabling models on Trainium.
- Lead cross-functional engineering efforts, influence multiple teams, and mentor senior engineers.
What they ask for
- 10+ years of non-internship professional software development experience.
- 7+ years with distributed training or RL systems at scale.
- Strong C/C++ and Python; experience with AI/ML model execution, inference serving, or compiler optimization.
- Technical leadership and mentorship of engineering teams.
Nice to have
- Deep hands-on expertise in at least one relevant area, such as distributed training, post-training and alignment, inference optimization, ML frameworks, or compiler/kernel optimization.
- Experience with adjacent areas such as EDA, video pipeline, GPU graphics, or HPC.
Pay
230,100.00 - 311,200.00 USD annually (base salary); 200,100.00 - 270,600.00 USD annually (base salary)
Tools
C/C++, Python, Trainium, Inferentia, PyTorch, JAX, TRL, veRL
AI focus
Optimize AI model pre-training, post-training and alignment (including RLHF/RLAIF and preference optimization), and inference serving on Trainium.
Model-written summary · the original posting is authoritative