Senior Forward Deployed Engineer - Model Customization and Optimization

About the role

As a Senior Forward Deployed Engineer on the Custom Model and Optimization team, you act as the technical lead embedded inside a strategic customer, responsible for building, optimizing and running custom generative AI models. You set the technical direction for model customization and optimization engagements, combining the depth of an AWS senior engineer with the pace of a startup builder, and you deliver measurable business outcomes in quality, latency and cost.

Key responsibilities

• Lead the technical engagement: embed with the customer as technical lead, partner with engineering and product leadership to define success metrics and the roadmap from proof of value to production, and communicate trade-offs and results to executives and engineers (up to 25% travel).
• Define the customization strategy: choose the path to quality, latency and cost targets (prompting, distillation, fine-tuning, RL, foundation models, adapters) and own the data strategy, synthetic data and evaluation methodology.
• Architect training and inference systems at scale: distributed training and RL post-training systems (PyTorch FSDP2, Megatron-LM, NeMo, verl) and production inference engines (SGLang, vLLM), plus RL harnesses (NeMo Gym, OpenEnv).
• Lead optimization on AWS AI chips: migrate workloads to AWS Trainium with the AWS Neuron SDK, diagnose gaps against GPU baselines and work with AWS Neuron and service teams to close capability gaps.
• Mentor a team of FDEs and scientists, run design and code reviews, take part in hiring, and share reusable fine-tuning recipes, evaluation frameworks and benchmarking tools across AWS.
• Own the delivery plan, track milestones and escalate risks early.

Required profile

• 5+ years of non-internship professional software development experience.
• 5+ years of programming with at least one language, and 5+ years leading design or architecture of systems.
• Experience as a mentor, tech lead or engineering team lead.
• Hands-on experience training, fine-tuning or optimizing deep learning models.
• Ability to explain technical concepts to technical and non-technical audiences, including senior leadership.
• Preferred: Bachelor's degree in computer science or equivalent, and 2+ years deploying ML models in customer production environments.

Required skills

• Preferred: experience with SGLang, vLLM, TensorRT-LLM, CUDA, Triton, Neuron SDK, SageMaker AI, SageMaker HyperPod, EKS.
• Preferred: experience with SFT, DPO, RLHF, RL with verifiable rewards or multi-turn agentic RL (Miles, NeMo-RL, OpenRLHF, SkyRL, TRL).
• Preferred: large-scale distributed training of LLMs (PyTorch FSDP2, Megatron-LM, DeepSpeed, NeuronX Distributed).

What we offer

• Base salary of $184,900-$250,200 in New York, plus sign-on payments and restricted stock units.
• Health insurance, 401(k) matching, paid time off and parental leave.