Forward Deployed Engineer (FDE) - AI Infra - SF

OpenAI•San Francisco, CA

1d•$185,000 - $300,000•Hybrid

About The Position

OpenAI’s Forward Deployed Engineering team partners with customers to turn research breakthroughs into production systems, working at the intersection of customer delivery and core platform development. Forward Deployed Engineers lead complex deployments of frontier models in production, embedding with customers to understand problems, structure delivery and ship solutions in urgent, ambiguous settings. The AI Infra Forward Deployed Engineer works directly with customer engineers, researchers and product engineers on optimization problems across semiconductors and AI infrastructure. The role involves writing and debugging software, running experiments against real workloads and hardware, and improving tools and products used across engagements. Work may include inference performance and chip-design optimization. The engineer defines problems and success criteria, builds software to test approaches, and carries work through validation and deployment. Success includes delivering results that meet correctness and performance checks and creating improvements reusable across engagements. The role owns delivery across workstreams, maintaining quality and pace while helping OpenAI learn through execution. It is based in San Francisco under a hybrid model requiring three days in the office per week. Travel of up to 50% is required, and relocation assistance is offered.

Requirements

At least 5 years of engineering or technical deployment experience, including customer-facing work.
Experience scoping and delivering complex systems in fast-moving or ambiguous environments.
Ability to write and review production-grade software and debug unfamiliar systems across code, integrations and runtime behavior.
Experience building or deploying systems powered by LLMs or generative models, with an understanding of how model behavior affects product experience.
Ability to simplify complexity and make fast, sound decisions under pressure.
Clear communication with engineers, product teams and customer stakeholders.
Ability to identify risks early and adjust without slowing delivery.
Calm judgment when stakes are high.
Strong product sense and contributions to productization efforts.
Hands-on experience with applied machine learning, model evaluation, data quality, experimentation or model-behavior analysis.
Ability to turn ambiguous technical problems into clear evaluations, credible baselines and reproducible experiments.
Comfort learning unfamiliar technical domains and working closely with researchers, product engineers and customer engineers.

Nice To Haves

Experience in semiconductor development, data-center systems or inference engineering.

Responsibilities

Own technical delivery for semiconductor and AI-infrastructure workstreams, from scoping and prototypes through validation, production deployment and adoption.
Work with customer engineers to understand workflows and constraints, define workloads, baselines and success criteria, and integrate solutions into customer workflows.
Scope and sequence delivery, identify dependencies and blockers early, and balance scope, speed and quality.
Build and debug production-quality software, including harnesses, integrations and product features applying frontier models to chip design, verification, inference and performance engineering.
Design reproducible experiments, evaluations, graders and benchmarks to test approaches and validate correctness and performance in real engineering environments.
Investigate model behavior and failure modes with researchers, contributing evaluation or training data and targeted experiments where useful.
Contribute to core product development, resolve tooling gaps with product engineers, ship improvements and bring field learning into product decisions.
Lead technical discussions and maintain alignment across customer, Product and Research teams; communicate capabilities, limitations and results and clarify ownership, risks and next steps.
Turn successful approaches into reusable software, tools and playbooks, reducing effort for future engagements and helping customer teams operate deployed solutions.