Software Engineer 5 – Model Runtime, AI Platform

About the Role: The Model Runtime team at Netflix owns the systems that train, align, and serve Netflix's most critical ML models. As a Software Engineer 5, you will thrive at the intersection of systems engineering and Machine Learning, shaping what Netflix can do with AI through highly scalable and differentiated ML infrastructure.

What You'll Do

• Build alignment and post-training infrastructure for reinforcement learning (GRPO, DPO, PPO) and preference optimization.
• Create infrastructure for multimodal and diffusion models, including distributed training and disaggregated serving.
• Engineer fault-tolerant training systems using FSDP and tensor/pipeline parallelism across clusters of hundreds of GPUs.
• Profile and tune from PyTorch operators down to GPU kernels to drive utilization improvements.
• Evaluate specialized accelerators and next-gen NVIDIA silicon to keep Netflix at the efficiency frontier.

What You Bring

• Deep experience in ML systems engineering, building infrastructure for training, fine-tuning, or inference of post-LLM era models at scale.
• Strong systems programming skills across the stack, down to GPU kernels and memory management.
• Hands-on experience with PyTorch internals, large-scale distributed training, and system-model codesign.
• Experience with cloud computing providers, preferably AWS, and a strong understanding of operations and observability.
• Bonus: Experience with vLLM, TensorRT, CUDA profiling, or contributing to open-source ML projects.

Benefits

• Top of market compensation range: $466,000 - $750,000 USD per year.
• Comprehensive health plans, mental health support, and family-forming benefits.
• 401(k) Retirement Plan with employer match and Stock Option Program.
• Flexible time off for salaried employees.
• Work in a unique culture with highly autonomous teams and outsized impact.