OpenAI

Software Engineer, RL Training Infra

San Francisco · 全职 · 算法与研究

岗位更新：2026-05-23资料抓取：2026-08-04查看原始招聘来源

需要对标的能力

大模型原理 · LLM 基础幻觉治理 · 输出可靠性Agent · 智能体设计上下文工程 · 记忆管理多 Agent 协作模型微调 · Post-TrainingAI Infra · 推理部署成本优化 · 性能调优评测 · Eval 方法论AI 工程设计模式

岗位职责与要求

ABOUT THE TEAM The Post-Training Frontiers team is responsible for training the frontier agents OpenAI ships to the world (GPT-Next). We train the flagship agentic models behind Codex, ChatGPT, and the API through large-scale reinforcement learning. The team’s work spans four areas. First, execution and science: working with teams across OpenAI to decide what can go into the final model and how, using scientific experiments and evals that are representative of the final pipeline so issues can be recognized early. Second, RL scaling: executing the final large-scale reinforcement learning run, making sure GPUs are used efficiently and training stays healthy. Third, research: improving horizontal capabilities like instruction following, factuality, memory, and multi-agent behavior, where the team’s broad visibility helps identify cross-cutting improvements across teams and domains. Fourth, engineering: maintaining the infrastructure stack and internal tools to ensure that both the final run and all integrations go as smoothly as possible and that the systems are easy to work with. ABOUT THE ROLE This role focuses on keeping our frontier RL training runs fast, reliable, and unblocked. You will work across engineering and infrastructure problems as they emerge, from scaling and orchestration issues to inference bottlenecks, numerical problems, and hardware failures, as well as supporting large horizontal integrations in the big run, like multi-agent capabilities or memory. This is a role for a strong generalist who quickly learns anything needed for the task, has high attention to detail, debugs deeply, and is motivated by fixing the highest-impact problem in front of the team. IN THIS ROLE, YOU WILL: - Keep large-scale async RL training runs moving by jumping into the most urgent engineering and infrastructure problems. - Debug issues across training systems, inference, orchestration, scaling, and distributed infrastructure. - Improve the reliability and efficiency of RL training runs. - Help researchers who are developing infrastructure-heavy integrations, such as multi-agent capabilities or memory. - Turn recurring operational issues into better tools, systems, processes, automations, or abstractions. - Work closely with research, infrastructure, and partner teams during tight model-run timelines. - Become useful quickly in messy, ambiguous areas where ownership matters more than a perfectly scoped project. - Debug hard failures in shipped or near-shipped models and turn messy qualitative behavior into concrete hypotheses, experiments, and fixes. YOU MIGHT THRIVE IN THIS ROLE IF YOU: - Are driven by having a large impact on the world and want to train and ship the best model in the world to our users. - Are a strong generalist engineer with experience in some layer of ML infrastructure. - Learn extremely quickly and are comfortable operating across unfamiliar layers. - Are highly independent and can plan and fix issues on your own when needed. - Have worked on RL, inference, scaling, training systems, orchestration, or adjacent infrastructure. - Are a strong debugger with high ownership, low ego, and excellent communication skills. - Can land in a messy area with tight timelines, become useful quickly, and gradually raise the quality of the whole system. - Thrive in fast-moving environments where reliability, speed, and judgment matter. - Are comfortable collaborating with multiple teams across OpenAI, both in research and in product. - Like building load-bearing systems and processes when that is what the team needs, even if the work is not glamorous. - Are very proactive about fixing any important issue that you encounter. - Are very good at prioritizing your work and often take an 80/20 approach, moving on to different problems after optimizing for low-hanging fruit. - Enjoy helping other people with their work. - Continuously try to automate the annoying parts of your work. NICE TO HAVE: - Experience supporting large-scale model training, async RL systems, or high-throughput ML infrastructure. - Experience debugging distributed systems across GPUs, networking, orchestration, or inference stacks. - A background in performance optimization, scaling, or production-critical infrastructure. - Experience directly supporting researchers or fast-moving model teams. To notify OpenAI that you believe this job posting is non-compliant, please submit a report through this form https://form.asana.com/?d=57018692298241&k=5MqR40fZd7jlxVUh5J-UeA. No response will be provided to inquiries unrelated to job posting compliance. At OpenAI, we believe artificial intelligence has the potential to help people solve immense global challenges, and we want the upside of AI to be widely shared. Join us in shaping the future of technology.

查看中文参考译文

关于团队 Post-Training Frontiers 团队负责训练 OpenAI 向世界推出的前沿 agent（GPT-Next）。我们通过大规模强化学习训练 Codex、ChatGPT 和 API 背后的旗舰 agentic 模型。团队的工作涵盖四个领域。第一，执行与科学：与 OpenAI 内部各团队合作，决定最终模型可以包含什么以及如何实现，使用能够代表最终 pipeline 的科学实验和 eval，以便尽早发现问题。第二，RL scaling：执行最终的大规模强化学习运行，确保 GPU 被高效利用且训练保持健康。第三，研究：提升 instruction following、factuality、memory 和 multi-agent behavior 等横向能力，团队广泛的视野有助于识别跨团队、跨领域的 cross-cutting 改进。第四，工程：维护基础设施栈和内部工具，确保最终运行和所有集成尽可能顺利，且系统易于使用。关于职位该职位专注于确保我们的前沿 RL 训练运行快速、可靠、不受阻碍。你将应对工程和基础设施中出现的各种问题，从 scaling 和 orchestration 问题到 inference 瓶颈、数值问题和硬件故障，同时支持大规模横向集成，如 multi-agent capabilities 或 memory。这是一个适合强 generalist 的角色，能够快速学习完成任务所需的任何知识，注重细节，深入 debug，并致力于解决团队面前最高影响的问题。在此职位中，你将： - 通过投身最紧迫的工程和基础设施问题，确保大规模 async RL 训练运行持续推进。 - 在 training systems、inference、orchestration、scaling 和分布式基础设施中 debug 问题。 - 提升 RL 训练运行的可靠性和效率。 - 协助研究人员开发基础设施密集型集成，如 multi-agent capabilities 或 memory。 - 将重复出现的 operational issues 转化为更好的工具、系统、流程、automations 或 abstractions。 - 在紧张的 model-run 时间线期间，与研究、基础设施和合作团队紧密协作。 - 在混乱、模糊的领域快速产生价值，在这些地方 ownership 比完美定义的项目范围更重要。 - 对已发布或即将发布的模型中的严重故障进行 debug，将混乱的定性行为转化为具体的假设、实验和修复。如果你具备以下条件，你可能适合此角色： - 渴望对世界产生重大影响，希望向用户训练并交付世界上最好的模型。 - 是一名强 generalist engineer，具备 ML infrastructure 某一层面的经验。 - 学习速度极快，能够舒适地在不熟悉的层面开展工作。 - 高度独立，能够在需要时自行规划和修复问题。 - 从事过 RL、inference、scaling、training systems、orchestration 或相关基础设施工作。 - 是一名强 debugger，具备高 ownership、低 ego 和出色的沟通能力。 - 能够在时间紧迫的混乱领域快速落地、快速产生价值，并逐步提升整个系统的质量。 - 在可靠性、速度和判断力至关重要的快节奏环境中蓬勃发展。 - 能够舒适地与 OpenAI 内部多个团队（包括研究和产品团队）协作。 - 乐于在团队需要时构建 load-bearing systems 和 processes，即使工作并不光鲜。 - 对遇到的任何重要问题都非常主动地修复。 - 非常擅长确定工作优先级，经常采用 80/20 方法，在优化完 low-hanging fruit 后转向不同问题。 - 乐于帮助他人完成工作。 - 持续尝试将工作中烦人的部分自动化。加分项： - 支持大规模模型训练、async RL systems 或高吞吐量 ML infrastructure 的经验。 - 在 GPU、networking、orchestration 或 inference stacks 上 debug 分布式系统的经验。 - performance optimization、scaling 或 production-critical infrastructure 方面的背景。 - 直接支持研究人员或快速移动的 model teams 的经验。如需通知 OpenAI 你认为此职位发布不合规，请通过此表单提交报告 https://form.asana.com/?d=57018692298241&k=5MqR40fZd7jlxVUh5J-UeA。与职位发布合规性无关的询问将不予回复。在 OpenAI，我们相信人工智能有潜力帮助人们解决巨大的全球挑战，我们希望 AI 的益处能够被广泛分享。加入我们，共同塑造技术的未来。

译文仅供辅助阅读，具体要求以原始招聘来源为准。