Senior AI Systems Engineer - LLM Inference & Performance

Fdetimes; Bellevue (WA); On-site; USD 180,000 - 280,000; Full time

Job summary

Snowflake is seeking AI-native researchers and systems developers to join the AI Research team and advance the state of the art in LLM inference systems and optimization. You will work across the full inference stack, including distributed serving, runtime systems, GPU kernels, and model-system co-design, driving latency, throughput, and cost improvements.

You will collaborate with model researchers and infrastructure teams to productionize innovations, publish findings, and push the boundaries

Qualifications

• Bachelor’s degree in computer science, electrical engineering, or related field; Master’s or PhD preferred.
• 5+ years in LLM inference systems, distributed AI systems, GPU systems, or HPC.
• Strong understanding of modern LLM inference architectures and tradeoffs for serving large-scale models.
• Hands-on experience with LLM inference and serving frameworks (vLLM, SGLang, TensorRT-LLM).
• Experience designing/optimizing inference runtimes: scheduling, batching, KV-cache, distributed execution, parallelism, disaggregated serving.
• Strong GPU architecture knowledge; experience with CUDA, Triton.
• Experience with performance-oriented libraries such as CUTLASS, cuBLAS, cuDNN.
• Experience profiling with Nsight Systems/Compute or similar tools.
• Ability to work independently, define questions, drive solutions amid ambiguity.
• Strong cross-team collaboration and communication.

Responsibilities

• Design and develop high-performance LLM inference systems across distributed serving, runtime systems, GPU execution, and kernels.
• Develop novel techniques to reduce latency, increase generation speed, throughput, and memory efficiency.
• Explore advanced inference techniques: speculative and parallel decoding, prefill/decode disaggregation, adaptive parallelism, and more.
• Build adaptive intelligent inference systems for new models, hardware, and workloads.
• Apply AI-native approaches to profiling, configuration search, and performance tuning.
• Identify high-impact performance issues and drive from research to production.
• Design distributed inference strategies across GPUs and nodes.
• Develop efficient multi-model serving, dynamic resource management, and workload-aware scheduling.
• Analyze and optimize GPU kernels for attention, MoE, and other components.
• Explore model-system co-design to unlock more efficient inference.
• Profile end-to-end workloads to identify bottlenecks across compute, memory, and networking.
• Collaborate with researchers, engineering, and product teams to deploy innovations.
• Open-source and publish innovations through blogs and conferences.

Skills

LLM inference systems

distributed AI systems

GPU systems

high-performance computing

profiling

model serving frameworks

experimental mindset

Education

Bachelor’s degree in Computer Science or Electrical Engineering

Master’s degree or PhD preferred

Tools

vLLM

SGLang

TensorRT-LLM

CUDA

Triton

CUTLASS

cuBLAS

cuDNN

Nsight Systems

Nsight Compute

Job description

Snowflake is seeking AI-native researchers and systems developers to join the AI Research team and advance the state of the art in LLM inference systems and optimization. You will work across the full inference stack, including distributed serving, runtime systems, GPU kernels, and model-system co-design, driving latency, throughput, and cost improvements.

You will collaborate with model researchers and infrastructure teams to productionize innovations, publish findings, and push the boundaries