Senior Solutions Architect, AI Inference
Core
Design and scale full-stack AI inference solutions for strategic partners, optimizing token generation on finite GPU resources and leading high-stakes technical conversations.
Role type
Senior Solutions Architect (AI Inference)
Builds
Reference architectures for disaggregated inference systems and optimized inference pipelines.
Domain
AI Infrastructure / GPU Computing / Cloud Partnerships
Deliverable
production ML models
Required skills
Full-stack inference design, Kubernetes cluster management, GPU resource optimization, Model optimization (TensorRT-LLM, vLLM, SGLang), Disaggregated serving, Multi-tier KV cache management, Speculative decoding, Quantization, Custom inference kernels, Agent design (sandboxing, memory, retrieval), Technical mentorship
Preferred skills
NVIDIA Dynamo, NIXL, Grove, Model customization (SFT, DPO, GRPO, RLVR), Open-source contributions
Technologies
NVIDIA Dynamo, Triton Inference Server, TensorRT-LLM, vLLM, SGLang, Kubernetes, TensorRT
Responsibilities
Work with inference partners to deploy distributed systems and bring back product insights; Build and operate inference recipes to improve efficiency; Accelerate inference pipelines using various backends; Evangelize devops best-practices for compute fabrics and observability; Provide mentorship and technical leadership for deployment and issue resolution
Seniority
Senior, hands-on IC with strategic partner leadership
