This is a fully remote position, open to applicants in California, +2 more states.

📋 Description

• Create, develop, and enhance agentic AI systems within the CUDA ecosystem, focusing on agent architectures, multi-agent workflows, tool usage, memory management, and orchestration.

• Construct data, environments, verification, reward, and evaluation systems to ensure continuous enhancement of model and agent capabilities.

• Collaborate in the design and optimization of agentic systems across NVIDIA's software and hardware infrastructure, from models and inference through compilers, runtimes, libraries, kernels, and GPUs.

• Work together with internal NVIDIA software, model, and hardware teams to integrate new functionalities into NVIDIA products.

⛳️ Requirements

• A Bachelor's degree in Computer Science, Electrical Engineering, or a related field, or equivalent experience; a Master's or PhD is preferred.

• Over 4 years of pertinent industry or academic experience in AI systems, machine learning, compilers, high-performance computing, or related disciplines.

• Practical experience in agent systems, coding agents, reinforcement learning, or contemporary AI inference systems.

• Proficient in C/C++, Rust, and Python programming languages.

• Strong foundation in software engineering principles.

• Experience with GPU programming and performance enhancement utilizing CUDA or similar accelerator platforms.

• Capacity to collaborate effectively across system boundaries.

• Proven history of developing impactful coding agents, autonomous software engineering systems, or developer tools.

• Practical experience optimizing and deploying with TRT-LLM, SGLang, vLLM, or Transformer Engine.

• Extensive knowledge of GPU systems and performance optimization.

• Publications or leadership in open source related to deep learning, agentic AI, reinforcement learning, compilers, high-performance computing, or AI systems.