Inference Performance Engineer - Latency & Cost

OpenAI; Los Angeles (CA); On-site; USD 295,000 - 555,000; Full time

Benefits offered by this job

Equity

Job summary

OpenAI is seeking a Software Engineer specializing in Inference Performance Optimization to work in Los Angeles. This role involves modeling inference performance across various layers, analyzing workloads, and collaborating with cross-functional teams for improvements. Responsibilities include building performance models, enhancing tooling for bottlenecks, and optimizing costs. The compensation ranges from $295K to $555K, plus equity. This is an opportunity to work on impactful optimizations within cutting-edge technologies.

Qualifications

• Enjoy reasoning from first principles about distributed systems and hardware efficiency.
• Comfortable working from application behavior to kernels and networking.
• Deep expertise in performance profiling, benchmarking, analysis, and optimization.

Responsibilities

• Build and refine performance models from microbenchmark results.
• Analyze inference workloads across applications and infrastructure.
• Enhance tooling to identify bottlenecks for latency and throughput.
• Partner with teams to turn insights into concrete improvements.

Skills

Distributed systems reasoning

Performance profiling

Benchmarking

Optimization

Collaboration

Job description

OpenAI is seeking a Software Engineer specializing in Inference Performance Optimization to work in Los Angeles. This role involves modeling inference performance across various layers, analyzing workloads, and collaborating with cross-functional teams for improvements. Responsibilities include building performance models, enhancing tooling for bottlenecks, and optimizing costs. The compensation ranges from $295K to $555K, plus equity. This is an opportunity to work on impactful optimizations within cutting-edge technologies.