Description
Join NVIDIA's PerfLab team as a Software Engineer Intern. Work with leading experts on groundbreaking projects in AI, computer graphics, and accelerated computing.
Responsibilities:
- Build and evaluate AI agent skills, MCP tools, harnesses, and agentic workflows to automate PerfLab operations and boost team efficiency.
- Build and implement data automation pipelines and infrastructure for performance test result collection, transformation, and visualization.
- Build interactive dashboards to present benchmark insights to internal and external customers.
- Work with NVIDIA's global experts to tackle intriguing challenges and drive impactful project developments.
Requirements:
- Pursuing a degree in Computer Science, Electrical Engineering, or a related field.
- Strong Python skills with experience in building CLI tools, REST APIs, or automation scripts in Unix and Windows environments.
- Hands-on experience building AI-powered applications such as AI agents, RAG pipelines, or MCP tools.
- Good understanding of ML/DL workloads (CNNs, Transformers, LLMs).
- Proficient in applying leading AI coding tools (e.g., Claude Code, Cursor, Codex) to accelerate daily work.
- Strong collaboration and communication skills, with the ability to learn quickly and work effectively in a dynamic environment.
- Excellent written and spoken English.
Nice to Have:
- Prior experience with agentic LLM frameworks like LangChain, LangGraph, or LlamaIndex.
- End-to-end experience developing Physical AI models, from data collection and preparation through training, evaluation, and deployment.
- Contributions to open-source projects in DevOps, infrastructure, or AI tooling.
- Demonstrated innovation and problem-solving through research, internships, hackathons, competitions, or personal projects.
- Experience writing technical blogs, giving tech talks, or creating demo videos to promote project outcomes.
NVIDIA offers highly competitive salaries and a comprehensive benefits package. Visit www.nvidiabenefits.com/ for more information.