Tech Stack
Tag name is followed by "@" symbol and proficiency level value.
About proficiency levels:
- 1-2 — basic awareness. Minimal hands-on experience, and a rudimentary understanding of the technology's purpose;
- 3-6 — daily use. Comfortable and regular usage, capable of handling common tasks and challenges related to the technology;
- 7-9 — you are an expert, you can teach others, you know all the pitfalls and tricks;
- 10 — exceptional knowledge, comprehensive understanding, and adeptness in all aspects of the technology, including advanced problem-solving. Think twice before claiming or demanding such level.
AI @ 7
CUDA @ 4
GPU @ 4
LLM @ 3
LangChain @ 3
Machine Learning
Networking @ 4
PyTorch @ 6
Python @ 4
Security @ 4
Spark @ 4
Technical Proficiency @ 6
TensorRT @ 3
vLLM @ 6
- 1-2 — basic awareness. Minimal hands-on experience, and a rudimentary understanding of the technology's purpose;
- 3-6 — daily use. Comfortable and regular usage, capable of handling common tasks and challenges related to the technology;
- 7-9 — you are an expert, you can teach others, you know all the pitfalls and tricks;
- 10 — exceptional knowledge, comprehensive understanding, and adeptness in all aspects of the technology, including advanced problem-solving. Think twice before claiming or demanding such level.
Details
Artificial intelligence is moving from passive assistance to agents that can reason, use tools, and complete work on a person's device. NVIDIA is building the software foundation that helps developers and partners deliver these experiences privately, efficiently, and responsibly across GeForce RTX, NVIDIA RTX PRO, RTX Spark, DGX Spark, and DGX Station systems.
The Senior Technical Program Manager will lead multifunctional initiatives involving local and hybrid AI agents, uniting engineering, research, product, security, developer relations, and external collaborators behind a clear roadmap and measurable results. The role covers local model inference, agent runtimes, Windows integration, developer tools, security controls, and user experiences. It also involves gathering insights from open ecosystems such as Hermes and OpenClaw, desktop-agent projects such as Perplexity, and domain-specific agents to guide NVIDIA's developer and user priorities.
Responsibilities
- Own the coordinated roadmap from prototype through release for local AI agent capabilities, including scope, achievements, owners, dependencies, and completion criteria.
- Translate product goals and ecosystem signals into harmonized plans covering agent frameworks, MCP and tool connections, memory and skills, computer use, local inference, model routing, and application integration.
- Partner with engineering and research teams to define evaluation and release gates for task success, latency, efficiency, memory use, power, reliability, setup time, privacy, and user trust.
- Coordinate programs across Windows platform groups, GPU and driver groups, model and runtime groups, product security, developer relations, Microsoft, OEMs, ISVs, and open-source communities.
- Build concise dashboards and decision forums that surface program health, technical trade-offs, risks, and evidence; communicate clearly with engineers and senior leaders.
- Use developer and customer feedback to improve onboarding, documentation, samples, compatibility, and adoption across supported NVIDIA systems.
Requirements
- Bachelor's degree in Computer Science, Computer Engineering, or a related field, or equivalent practical experience.
- 8+ years leading complex software, platform, systems, or AI/ML programs from concept through production release.
- Technical proficiency in contemporary AI software stacks, including model prediction, agent coordination, tool application, evaluation, and local or hybrid deployment.
- Experience building coordinated plans across multiple engineering organizations and resolving technical dependencies without direct authority.
- Experience defining measurable quality and release criteria, using data to make trade-offs, and separating prototypes from validated capabilities.
- Ability to explain architecture, risk, and program status clearly to technical teams, partners, and executives.
Preferred Qualifications
- Hands-on experience with Python, C++, or software automation.
- Familiarity with OpenClaw, Hermes, LangChain or similar agent frameworks; MCP; and persistent memory, skills, multi-agent, or computer-use patterns.
- Familiarity with local inference technologies such as TensorRT-LLM, Ollama, llama.cpp, vLLM, PyTorch, ONNX Runtime, or Windows ML, as well as model optimization or quantization.
- Experience with Windows systems, CUDA or GPU acceleration, sandboxing, policy controls, privacy-aware networking, or secure credential handling.
- Experience collaborating with Microsoft, OEMs, ISVs, open-source maintainers, or domain-solution partners.
Compensation and Benefits
The base salary range is USD 168,000–258,750 for Level 4 and USD 200,000–322,000 for Level 5. The position is also eligible for equity and benefits. NVIDIA offers a comprehensive benefits package.
Applications will be accepted at least until August 28, 2026. This posting is for an existing vacancy. NVIDIA uses AI tools in its recruiting processes and is an equal opportunity employer.