Performance Engineer (AI Inference)

TL;DR

Performance Engineer (AI Inference): Developing and optimizing high-throughput inference systems for Claude with an accent on throughput, latency, reliability, and correctness. Focus on cross-layer performance investigations, building observability tools, and bridging the gap between actual fleet performance and theoretical rooflines.

Location: Hybrid; must be based in or attend one of the offices (San Francisco, New York City, or Seattle) at least 25% of the time.

Salary: $350,000 - $850,000 USD

Company

hirify.global is a public benefit corporation focused on creating reliable, interpretable, and steerable AI systems for the benefit of society.

What you will do

• Conduct cross-layer performance investigations to identify root causes for gaps in throughput, latency, and reliability.
• Own and improve the correctness evaluation pipeline to validate model output quality across hardware platforms and serving configurations.
• Develop observability dashboards and modeling tools to make system interactions legible across the stack.
• Partner with kernel, serving, routing, and capacity teams to implement high-impact optimizations.
• Prioritize and stack-rank a large surface area of optimization opportunities based on impact and effort.

Requirements

• Hands-on experience in performance engineering, including profiling, roofline analysis, and root-cause investigation in production systems.
• Proficiency in Python and the ability to instrument and contribute to large production codebases.
• Strong data analysis skills using SQL, pandas, or similar tools.
• Ability to communicate quantitative results clearly to influence priorities across teams.
• Genuine interest in correctness as an engineering discipline, including numerics and regression detection.
• Must be based in or able to attend the US offices (SF, NYC, or Seattle) at least 25% of the time.

Nice to have

• Experience with ML systems, specifically training or inference infrastructure and LLM serving stacks.
• Familiarity with GPU/TPU/accelerator performance concepts such as memory bandwidth and quantization.
• Reliability engineering experience for high-throughput services, including autoscaling and load balancing.
• Experience building observability or telemetry for distributed systems.
• Experience with model evaluation or numerical regression-detection pipelines.

Culture & Benefits

• Collaborative research environment based on the "big science" approach.
• Competitive compensation and optional equity donation matching.
• Generous vacation and parental leave.
• Flexible working hours and high-quality collaborative office spaces.
• Visa sponsorship available for qualified candidates.