Systems Generalist, GPT Infrastructure

About the role

This role focuses on developing an automated platform for optimizing AI inference. You will build systems that generate, compile, execute, and evaluate candidate kernels and configurations across various hardware. The work involves creating both OpenAI-hosted control planes and partner-side software for performance verification. As part of the GPT Infrastructure team in San Francisco, you will sit at the core of the services that make model inference fast and cost-efficient. The mission is to treat optimization as an automated, continuous process rather than a set of manual experiments: pipelines that repeatedly test variations in stack kernels, runtimes, and hardware configurations, then qualify the winners for production deployment. Because the work spans OpenAI-owned and partner-owned compute environments, the systems you build must be secure, observable, and designed for durable operations. This is a senior engineering role for someone comfortable owning significant subsystems end to end, designing reliable APIs and workflows, and scaling work across the organizational and technical boundaries separating OpenAI systems from partner environments.

Key facts

What you'll do

• Design and operate APIs and control-plane services for extended optimization campaigns, including scheduling, retries, and observability.
• Build secure partner-side software to compile, execute, verify, and benchmark artifacts on third-party accelerator hardware.
• Integrate hardware profiles, toolchains, compilers, runtimes, and inference engines into a single, repeatable optimization loop.
• Convert research prototypes into reliable product surfaces with clear contracts and dependable results.
• Implement correctness and performance evaluation for latency, throughput, memory usage, and cost efficiency.
• Build artifact and qualification workflows for optimized kernels, binaries, and reports.
• Collaborate with research, engineering, infrastructure, security, product, and strategic partners to deliver production solutions.
• Drive technical design and execution for cross-functional initiatives spanning OpenAI and the rest of the ecosystem.

Requirements

• 8+ years of professional software engineering experience in large-scale distributed systems, infrastructure platforms, or cloud services.
• Strong programming foundation in C++, Python, Go, or Rust.
• Experience designing and operating highly available systems: APIs, jobs, or durable workflows that run in production.
• Deep grounding in distributed systems, Linux, networking, storage, containers, and modern cloud architectures.
• Experience debugging complex systems using measurement, profiling, and principled benchmark analysis.
• The proven ability to lead complex technical initiatives and operate across teams and boundaries.

Nice to have

• Experience with AI infrastructure, inference-serving systems, or large-scale ML training/automation at service scale.
• Compilers, runtimes, kernels, or performance engineering background (LLVM, MLIR, Triton, CUDA, ROCm).
• Familiarity with accelerators, hardware, architecture, or vendor toolchains.
• Experience with inference serving frameworks like vLLM, SGLang, or Triton Inference Server.
• Developer platforms, external APIs, remote execution, or secure partner-facing systems.
• Experience working with cloud, hardware, or partner organizations.

Skills & tools

• C++ / Python / Go / Rust
• Linux systems engineering
• LLVM, MLIR, Triton, CUDA, ROCm
• vLLM, SGLang, Triton Inference Server
• GPU and accelerator hardware
• Distributed systems and reliability

Practical notes

a hybrid role based in San Francisco. OpenAI is committed to building and ensuring that general-purpose AI benefits all of humanity and is an equal opportunity employer. Background checks are performed in accordance with applicable law, and qualified candidates with an arrest or conviction record will be considered consistent with laws such as the San Francisco Fair Chance Ordinance. Reasonable accommodations for applicants with disabilities are available on request. The work here combines hands-on systems engineering with close partnership across the AI ecosystem, and it directly influences how efficiently OpenAI's flagship models are served.

About the company

OpenAI is an AI research and deployment company dedicated to ensuring that general-purpose artificial intelligence benefits all of humanity. We push the boundaries of the capabilities of AI systems and seek to safely deploy them to the world through our products.