OpenAI Applied AI
Machine Learning Engineer, API Multicloud
- Location San Francisco
- Compensation 223,000 – 423,000 USD / year
- Type FullTime
- Team Core Product & Platform
- Seniority mid
- Posted 2026-08-13
Original posting ↗ You apply on the company site — we never collect applications.
Extracted automatically from public job postings. Always verify details on the original posting before applying.
role details
Partner with customers and internal teams to define model behaviors, diagnose issues, and improve training and system requirements.
Summary generated by AI from the original posting.
Hard requirements to check first
- Clearance：not mentioned in the posting
- Work auth：not mentioned in the posting
- Travel：customer-facing; travel not stated
Skills
PyTorchTensorFlowAWSFineTuningAgentsMachine LearningPost-training workflowsData pipelinesModel behaviorAPI/infrastructure integrationCloud-native environmentsProduction ML systems
Excerpt from the original posting
About the Team OpenAI’s API Multicloud team is responsible for extending OpenAI’s API platform into strategic cloud environments, starting with AWS . The team’s mission is to distribute OpenAI’s API broadly and safely by enabling key API technologies in AWS-native environments, in close partnership with Amazon and internal teams across Codex, Research, Safety Systems, and Applied. The team is focused on bringing core developer and enterprise capabilities into cloud-native environments, including AWS-hosted Codex, model customization / post-training as a service, and new stateful runtime environments for agentic workloads. This work sits at the intersection of production ML systems, developer platforms, model behavior, and large-scale infrastructure. About the Role We’re hiring Machine Learning Engineers to build and improve the AI systems that help strategic partners adapt OpenAI models to important use cases in cloud-native environments. This role spans post-training workflows, evaluation, data pipelines, model behavior, and API/infrastructure integration. You’ll work at the boundary between partner needs and core ML systems: helping teams understand what is and isn’t working, diagnosing issues in training and evaluation workflows, and turning those learnings into improvements to the underlying platform. You should enjoy working with external technical partners, extracting the real goal from messy requests, and pushing back or reframing when the requested experiment is not the highest-leverage path. You’ll collaborate closely with Research, Applied, Safety Systems, infrastructure teams, and external technical partners to solve ambiguous model-performance problems. When you succeed, strategic partners and internal teams will be able to improve model behavior with confidence, driving measurable product improvements while the systems behind that work become more reliable, scalable, and effective over time. In this role, you will - Partner with strategic customers and internal teams to define target model behaviors, diagnose failure modes, and translate real-world needs into training, evaluation, and system requirements. - Build and scale production ML systems for model customization, post-training, and fine-tuning-as-a-service workflows. - Design, run, and interpret experiments to see whether training and customization workflows are producing the intended outcomes, and identify changes to data, evaluation, training, or infrastructure that improve perf…
Verify before you apply
Before you apply
- 157 FDE roles in San Francisco — every role we track in this city, refreshed daily
- What these roles pay — 223 postings with published pay — midpoint median $231,000
- Market report — how all 922 live roles break down, with sample size and method
- Career path — what changes in scope between FDE levels