Senior AI Systems Architect: Scalable Agent Runtimes

NVIDIA Corporation; United States; Remote; USD 224,000 - 431,000; Full time

Job summary

NVIDIA Corporation is hiring a Senior Software Architect in the Agent Harness & Runtime Engineering team in the United States. The role focuses on building foundational, scalable systems for agentic AI across runtimes, inference, evaluation, and orchestration.

You will address large-scale distributed challenges, optimize performance, and lead across research, engineering, product, and infrastructure teams with hands-on coding in Python/C++/Go/Rust.

Qualifications

• BS, MS, or equivalent experience in CS/AI with 12+ years in industry.
• Strong foundation in modern AI incl. LLMs, multimodal models, inference, agentic AI, and evaluation.
• Deep expertise in software architecture, distributed systems, and large-scale design.
• Hands-on programming in Python, C++, Go, or Rust with production-quality code.
• Experience building large-scale systems with distributed execution, data processing/HPC/cloud or GPU infra.
• Ability to reason across stack from model behavior to runtime and infrastructure.
• Proven track record delivering ambiguous technical problems end-to-end.

Responsibilities

• Architect and build scalable systems for agentic AI across runtimes, inference, evaluation, and orchestration.
• Solve large-scale distributed execution, data pipelines, HPC, cloud, Kubernetes, and GPU environments.
• Optimize reliability, scalability, performance, resources, and developer experience.
• Prototype ideas, write production-grade code, and provide technical leadership across teams.

Skills

Python

C++

Go

Rust

Education

BS, MS or equivalent in CS/related field

Tools

Kubernetes

Job description

NVIDIA Corporation is hiring a Senior Software Architect in the Agent Harness & Runtime Engineering team in the United States. The role focuses on building foundational, scalable systems for agentic AI across runtimes, inference, evaluation, and orchestration.

You will address large-scale distributed challenges, optimize performance, and lead across research, engineering, product, and infrastructure teams with hands-on coding in Python/C++/Go/Rust.