Senior System Software Engineer – Dynamo Tools

NVIDIA•Us, CA

About The Position

We are seeking a Senior System Software Engineer to own and advance the AI-Perf analysis, NVIDIA’s flagship framework for benchmarking, experimentation, and analysis of LLMs, Generative AI, and deep learning inference workloads. In this role, you’ll combine systems research, distributed systems engineering, and applied AI, enabling reproducible performance evaluation, influencing internal platforms, and providing tooling that empowers researchers and engineers globally.

Requirements

Bachelor’s, Master’s, or PhD in Computer Science, Computer Engineering, or related field—or equivalent experience.
3+ years of experience in systems software, distributed performance engineering, or AI infrastructure research.
Expert-level Python skills, including profiling, optimization, automation, and debugging of complex systems.
Deep knowledge of distributed systems concepts, including scalability, concurrency, fault tolerance, and performance trade-offs.

Nice To Haves

Experience designing or maintaining performance benchmarking frameworks or tooling for AI/ML systems.
Hands-on experience with LLMs and deep learning frameworks such as PyTorch, TensorFlow, TensorRT, or ONNX Runtime.
Contributions to open-source or research projects in AI performance, infrastructure, or distributed systems.
Experience running large-scale inference experiments across cloud and on-prem environments (AWS, Azure, GCP, bare metal).

Responsibilities

Lead the design, development, and roadmap of AI-Perf, defining benchmarking methodologies, performance metrics, and reproducible experimental workflows.
Build scalable and high-performance features to measure latency, throughput, and efficiency across AI models and distributed systems.
Partner with AI researchers, platform teams, and engineers to translate experimental challenges into robust, user-friendly performance tooling.
Integrate AI-Perf with the Dynamo Inference Stack, other NVIDIA inference stacks, and open-source inference frameworks, delivering end-to-end performance insights for researchers and production users.