Harness and Platform Engineer, AI Safety and Security Engineering
据了解，NVIDIA 是世界上最受欢迎的雇主之一，一些人为我们工作！我们相信开放型号是美国人工智能领导和网络安全的基础，并且通过参与，透明度和广泛科学审查来对人工智能的信任增长。我们的 AI 安全与安全工程团队构建并评估使用人工智能驱动工具，帮助发现，验证和补丁软件漏洞。代理带和运行的平台是其他一切建立在基层上的基础。我们正在寻找一个杆/平台工程师来建立这个基础。你将在使程序的工作成为可能，可重复和可信的过程中发挥关键作用。当一个研究人员进行实验时，你的平台决定结果是否可以明天再现。我们的目前方向是基于 NVIDIA NeMo 代理工具，你会帮助塑造它去哪里。你将保持整个循环可见，从环境到运行到结果。你将与安全研究和评估工程师密切合作，学习他们的工作流程并平滑他们的路径。好的平台工作看起来像不太多惊喜：行为环境，保持版本的工具，并且重复运行。当某种东西破裂时，你会与最先感受到的人进行调试。你将研究人员时间视为宝贵，并在任何地方消除摩擦。你会做什么：· 开发利用：设计，实施和维护代理使用。·评估基础设施：构建我们运行和复制实验所用的系统。· 可重现性：自己的环境，工具和可重复运行。· 与安全和评估工程师合作。我们需要看到的是什么：· 学士学位 （或同等经验）5年以上软件工程经验。·技术核心：强大的 Python 工程加上容器化环境和 CI/CD 的舒适性。·基础设施熟悉：经验与代理框架，LLM 管弦，ML 基础架构或评估杆。·复制性的本能：关注版本工具和可重复结果。如何脱而出群众：·NeMo 体验：使用 NVIDIA NeMo 或类似代理框架进行实践工作。·安全背景：接触到安全工具或漏洞研究。·开发人员经验：建立内部工具其他工程师喜欢使用。你的基本工资将根据你的情况确定
查看英文原文
NVIDIA is widely recognized as one of the most desirable employers, with some of the most talented people in the world working for us! We believe open-weight models are foundational to American AI leadership and cybersecurity, and that trust in AI grows through participation, transparency, and broad scientific scrutiny. Our AI Safety & Security Engineering team builds and evaluates AI-powered tooling that helps find, validate, and patch software vulnerabilities. The agent harness and the platform it runs on are the substrate everything else builds on.
We are looking for a Harness/Platform Engineer to build that foundation. You will play a critical role in making the program's work possible, repeatable, and trustworthy. When a researcher runs an experiment, your platform decides whether the result can be reproduced tomorrow. Our current direction builds on NVIDIA NeMo agent tooling, and you will help shape where it goes. You will keep the whole loop visible, from environment to run to result. You will work closely with security research and evaluation engineers, learning their workflows and smoothing their paths. Good platform work here looks like fewer surprises: environments that behave, tooling that stays versioned, and runs that repeat. When something breaks, you will debug it with the people who felt it first. You will treat researcher time as precious and remove friction wherever you find it.
What You'll Be Doing:
- Harness development:Design, implement, and maintain the agent harness.
- Evaluation infrastructure:Build the systems we use to run and reproduce experiments.
- Reproducibility:Own environments, tooling, and repeatable runs.
- Collaboration:Partner with security and evaluation engineers on their workflows.
What We Need To See:
- Bachelor's degree (or equivalent experience) with 5+ years of software engineering experience.
- Technical core:Strong Python engineering plus comfort with containerized environments and CI/CD.
- Infrastructure familiarity:Experience with agent frameworks, LLM orchestration, ML infrastructure, or evaluation harnesses.
- Reproducibility instincts:Care for versioned tooling and repeatable results.
Ways to Stand Out from the Crowd:
- NeMo experience:Hands-on work with NVIDIA NeMo or similar agent frameworks.
- Security context:Exposure to security tooling or vulnerability research.
- Developer experience:Building internal tools other engineers enjoy using.
Your base salary will be determined based on your location, experience, and the pay of employees in similar positions. The base salary range is 152,000 USD - 241,500 USD.You will also be eligible for equity and benefits.
Applications for this job will be accepted at least until August 15, 2026.This posting is for an existing vacancy.
NVIDIA uses AI tools in its recruiting processes.
NVIDIA is committed to fostering an inclusive work environment and proud to be an equal opportunity employer. As we highly value diversity in our current and future employees, we do not discriminate (including in our hiring and promotion practices) on the basis of race, religion, color, national origin, gender, gender expression, sexual orientation, age, marital status, veteran status, disability status or any other characteristic protected by law.Originally posted on Himalayas