Product Manager, AI Platform SW - Agentic AI Kernel Generation

NVIDIA•Santa Clara, CA

8d•$168,000 - $327,750

About The Position

NVIDIA's AI Software Platforms team is building agentic AI infrastructure that enables coding agents to synthesize, optimize, and deploy GPU kernels automatically. This role focuses on products that connect data pipelines, evaluation suites, and GPU-accelerated runtimes, helping developers release faster and better-performing inference and training solutions. The Product Manager will champion AI agents and LLM-based coding workflows that generate optimized kernels. Working closely with engineering, research, customers, NVIDIA leaders, and ecosystem partners, the role defines strategy and roadmaps for products spanning the agent lifecycle, from data collection and synthetic data generation through evaluation, deployment, and continuous improvement. The products support coding agents that generate, refactor, and optimize CUDA kernels and graph-level execution plans across GPU architectures.

Requirements

7+ years of technical product management or closely related experience shipping developer or platform products in AI, ML infrastructure, or high-performance computing.
End-to-end ownership and demonstrated impact.
Experience in the AI agent or LLM space, including developing or productizing coding agents.
Experience with multi-agent orchestration and self-healing or code loops that improve over time.
Experience connecting agents to compilers or execution environments.
A proven record of creating and releasing automated testing or evaluation suites measuring agents with objective metrics such as correctness, performance, and latency.
BS or MS in Computer Engineering, Computer Science, or a related technical field, or equivalent experience in parallel computing architectures and systems.

Nice To Haves

PhD or equivalent experience in Computer Engineering, Computer Science, or another technical specialty.
Experience building or launching coding-agent platforms or copilots used by development teams at scale.
Contributions to performance-critical open-source projects, such as Triton, TVM, FlashAttention, kernel libraries, or agent frameworks, with clear community adoption and impact.
Research experience in GPU kernel optimization, collective or group communication algorithms, multi-agent systems, or ML model serving and inference architectures, demonstrating end-to-end systems thinking.
Experience crafting cost-per-inference or cost-per-token models that incorporate hardware utilization, energy efficiency, and cluster scaling, and using those models to guide product strategy and tradeoffs.

Responsibilities

Architect agent-focused products that enable coding agents to generate, refactor, and optimize CUDA kernels and graph-level execution plans across diverse GPU architectures.
Define the end-to-end data lifecycle for agent training and evaluation, including dataset curation, artificial data creation, and benchmark suites for correctness, latency, and adaptability.
Partner with CUDA, kernel, and compiler engineering teams to safely and observably integrate agents with compilers, profilers, execution sandboxes, and runtimes.
Collaborate with internal and external developers, NVIDIA leaders, and ecosystem partners to drive multi-agent orchestration, prioritize features, and deliver launches and messaging for agentic AI kernel generation.