Nvidia AI-Accelerated System Software Engineer 

United States, California 

861304074

What you will be doing:

Design, develop and deliver core components of our next-generation engineer productivity platforms.

• Envision, build, deploy and improve highly scalable systems at the speed of thought

• Disambiguate high-level requirements into actionable Agentic AI tasks with clear vetting and ownership of output

• Jump between microservices and monoliths, on-prem and cloud-compute as needed to build efficient products

• Convert legacy codebases into modern powerhouses with parallelized AI migration

• Collaborate with engineering teams to identify and alleviate bottlenecks in their daily tasks

• Be adept and flexible to changing priorities and updating your codebases as business needs evolve

What we need to see:

Proficiency in thought-to-code and rapid understanding of existing or AI-generated code to debug and identify pathways to improve performance, resilience and observability.

• M.S. in EE/CS (or equivalent experience)

• Minimum 5 years developing large-scale software applications in an enterprise engineering environment using Python

• Internalized computer science fundamentals in algorithms/datastructures/complexityanalyses

• Rich history of wrangling LLMs (GPT 4o through Claude Opus 4.x) to ship real software

• Deep, demonstrable intuition of the capabilities and pitfalls of LLM code generation ("vibe coding")

• Track record of shipping and supporting AI-generated code into enterprise engineering applications

Ways to stand out from the crowd:

Core system-engineering skills including the ability to visualize strongly-coupled systems from architecture down to OS and compute.

• Solid fundamentals of UNIX internals, filesystems, job-scheduling, processes, synchronization, and locks

• Experience with Perforce, NFS, LSF and running services on managed infrastructure without root/sudo

• Experience building maintainable software using off-the-shelf tools and open-source libraries to solve feature gaps quickly

• You've handled your own deployment and infrastructure on a cloud provider and on bare-metal servers

• Rapid problem decomposition and clear communication when delegating work to AI Agents