CareerPlan

Member of Technical Staff - GPU Inference Engineer

San FranciscoFull-time2026-11-19 → 2026-11-22
Hybrid

Core

Optimizing GPU inference stacks (vLLM, SGLang) to improve speed and cost for LLM evaluation, synthetic data generation, and API serving.

Role type

Senior IC GPU Inference Engineer

Builds

GPU serving stack, inference pipelines, and API service performance

Domain

AI/ML Infrastructure, GPU Computing, LLM Serving

Deliverable

production ML models

Required skills

LLM inference optimization, vLLM, SGLang, GPU fundamentals, profiling and benchmarking, Python, PyTorch, open-source codebase navigation

Preferred skills

CUDA/Triton kernel development, FlashInfer contributions, high-throughput batch inference

Technologies

vLLM, SGLang, PyTorch, Hugging Face Transformers, CUDA, Triton

Responsibilities

Optimize GPU serving stack configuration and scheduling; Profile and benchmark inference engines; Integrate foundation models into serving frameworks; Diagnose throughput and latency issues; Design inference pipelines for evaluation and data generation

Seniority

Senior, hands-on IC