NVIDIA
Senior Staff Software Engineer - AI Agent Platform
$200K–$391K
US, CA, Santa Clara · Senior Staff · Full-time · Posted Oct 16, 2026
- Experience
- 12+ years in software engineering.
- Education
- Bachelor's or Master's degree in Computer Science, Engineering, or related field (or equivalent experience).
About the job
Designs and operates the infrastructure and platform services that let teams build, deploy, orchestrate, and run AI agents at production scale.
What you'll do
- Build services for the full agent lifecycle and multi-tenant Kubernetes execution environments.
- Develop CI/CD pipelines, agent-framework integrations, messaging and data pipelines, and session management.
- Implement authentication and secrets management, and partner with security teams on compliance and access controls.
What they ask for
- 12+ years in software engineering; ideally in platform engineering, infrastructure, or developer tools.
- Experience building and scaling production AI agents using frameworks such as Claude Code, Codex, or LangGraph.
- Deep Kubernetes expertise; strong Python and production API experience; distributed-systems experience with Kafka, Redis, and MongoDB or PostgreSQL.
- Experience with GitLab CI and ArgoCD, AI data platform components and developer-facing APIs, and authentication and secrets management.
Nice to have
- Experience with agent platforms or protocols such as MCP, A2A, LangChain, or LangGraph; RAG, embedding pipelines, or vector databases.
- React or Vue experience, or contributions to open-source infrastructure or platform tooling.
Pay
200,000 USD - 322,000 USD base salary (Level 5); 248,000 USD - 391,000 USD base salary (Level 6)
Tools
Kubernetes, Python, FastAPI, Flask, Kafka, Redis, MongoDB, PostgreSQL
AI focus
Builds platform infrastructure for developing, deploying, orchestrating, and operating autonomous AI agents, including agent lifecycle services, framework integrations, knowledge and context data systems, and session persistence.
Model-written summary · the original posting is authoritative