AI-Native Systems Engineer, TensorRT — Hybrid & Equity

NVIDIA Corporation; Santa Clara (CA); On-site; USD 152,000 - 288,000; Full time

Benefits offered by this job

Equity

Benefits

Hybrid work model

Job summary

NVIDIA Corporation in Santa Clara, CA is seeking a Senior AI-Native Systems Software Engineer for the TensorRT team. The role focuses on building an AI-native framework that scales with many AI agents, delivering high-performance inference across LLMs, diffusion models, and multimodal systems.

You will design scalable architectures, prototype with state-of-the-art models, and optimize end-to-end performance from Python orchestration to engine-level CUDA/C++ code.

Qualifications

• 4+ years of software development experience.
• Strong modern C++ skills (C++11/14/17+).
• Experience with deep learning inference frameworks.
• Ability to translate customer needs into technical requirements and scalable solutions.

Responsibilities

• Architect an AI-native framework that scales beyond human capacity with multi-agent parallel testing.
• Advance agentic workflows by building AI-native tools and orchestrators for large-scale codebases.
• Prototype with state-of-the-art models and dispatch AI swarms to integrate capabilities.
• Ensure a high-performance user experience from model families to production deployment.
• Optimize performance at the intersection of Python orchestration and C++ engine code.

Skills

Modern C++

Deep learning

System design

End-to-end product sense

Pragmatic execution

Collaborative mindset

Education

BS/MS/PhD in CS/CE/AI or equivalent

Tools

CUDA

Job description

NVIDIA Corporation in Santa Clara, CA is seeking a Senior AI-Native Systems Software Engineer for the TensorRT team. The role focuses on building an AI-native framework that scales with many AI agents, delivering high-performance inference across LLMs, diffusion models, and multimodal systems.

You will design scalable architectures, prototype with state-of-the-art models, and optimize end-to-end performance from Python orchestration to engine-level CUDA/C++ code.