Performance Engineer (AI)

Location: Full-time onsite in Austin, Texas or Sunnyvale, California

Salary: $210K–$250K plus equity

Company

hirify.global is an AI hardware startup developing silicon photonics and programmable metasurface technology for energy-efficient optical inference accelerators.

What you will do

• Own the performance and energy metrics used for architecture, product, and leadership decisions.
• Produce consistent measurements across roofline and limiter models, architecture performance models, RTL simulation, and competitor hardware.
• Define and maintain workload methodology across model, sequence length, batch, precision, inference phase, and parallelism.
• Bring up inference workloads from Hugging Face, PyTorch, research papers, and serving stacks including vLLM, SGLang, TensorRT-LLM, and Triton Inference Server.
• Measure competing GPUs and accelerators end to end, including cloud or lab environments, drivers, images, and run recipes.
• Report TTFT, inter-token latency, throughput, throughput per watt, and energy, with reproducible harnesses, configurations, plots, and logs.

Requirements

• BS or MS in Computer Engineering, Electrical Engineering, Computer Science, or equivalent practical experience.
• 5+ years of experience in GPU performance engineering, accelerator benchmarking, HPC performance measurement, or ML systems measurement.
• Experience building or operating benchmark harnesses that generate measured results on real GPUs or accelerators.
• Hands-on experience with roofline analysis, limiter analysis, analytical performance modeling, and GPU profiling using NVIDIA Nsight Systems, Nsight Compute, or equivalent tools.
• Working knowledge of LLM inference stacks, including prefill versus decode, continuous batching, and Mixture-of-Experts models.
• Proficiency in Python and Linux, plus experience with cloud GPU operations, containers, instance types, drivers, quotas, and cost management.

Nice to have

• Experience correlating performance models or RTL/Verilator simulations with measured silicon or GPUs.
• GPU kernel experience with CUDA, CUTLASS, or Triton, or familiarity with PyTorch internals.
• Knowledge of PagedAttention, FlashAttention, speculative decoding, and disaggregated prefill.
• Distributed inference experience with collectives, all-reduce, NCCL, NVLink, InfiniBand, MLPerf, or production benchmarking pipelines.
• Background at a hyperscaler, GPU vendor, accelerator company, or inference lab.

Culture & Benefits

• Collaborative engineering environment focused on photonics, AI, and computational performance.
• 100% coverage of base health plan premiums for employees and dependents, plus HSA contributions.
• Unlimited PTO.
• 401(k) matching and stock option opportunities.
• Dental, vision, life, hospital, critical illness, and accident insurance, with personalized plan options.