Inference Performance Engineer: Cost & Capacity Modeling

Visa Hunt; San Francisco, Northern (CA, KY); Hybrid; USD 150,000 - 190,000; Full time

Job summary

OpenAI is seeking a role focused on modeling inference performance across application, model, and fleet layers. You will build cost-to-serve estimates from microbenchmarks and create tools that help teams reason about latency, capacity, utilization, and cost tradeoffs.

The role emphasizes deep expertise in profiling, benchmarking, and optimization, with collaboration across engineering and research to implement concrete improvements in production systems.

Qualifications

• Deep expertise with performance profiling, benchmarking, analysis, and optimization.

Responsibilities

• Build and refine performance models for cost-to-serve estimates.
• Analyze inference workloads end to end across applications, models, and fleet infrastructure.
• Collaborate with engineering and research teams to translate insights into improvements.

Skills

Performance profiling

Benchmarking

Analysis

Optimization

Distributed systems

Model inference

Job description

OpenAI is seeking a role focused on modeling inference performance across application, model, and fleet layers. You will build cost-to-serve estimates from microbenchmarks and create tools that help teams reason about latency, capacity, utilization, and cost tradeoffs.

The role emphasizes deep expertise in profiling, benchmarking, and optimization, with collaboration across engineering and research to implement concrete improvements in production systems.