Principal Software Engineer — Agentic AI Applications and Foundations
Core
Architecting and hardening production-grade infrastructure for NVIDIA's internal Enterprise AI agents, ensuring reliability, scalability, and operational excellence across desktop, web, and service-based products.
Role type
Principal Software Engineer (Agentic AI Infrastructure)
Builds
Resilient frontends, backend APIs, distributed services, data flows, deployment systems, and shared agent capabilities (orchestration, memory, evaluation frameworks) for NVIDIA's internal AI tools.
Domain
Enterprise AI, Agentic Workflows, Full-Stack Engineering
Deliverable
production ML models | product features | infrastructure
Required skills
Full-stack architecture (frontend/backend/platform), distributed systems design, release engineering, observability, incident response, testing strategy, cross-functional leadership, AI application patterns (LLM, RAG, tool use, agentic workflows), GPU-accelerated systems.
Preferred skills
Desktop/client application hardening at scale, engineering velocity improvement via frameworks, reusable infrastructure for AI products, identity/discovery/trust systems, NVIDIA AI technologies (NeMo, NIM, Nemotron, TensorRT-LLM).
Technologies
TypeScript, JavaScript, React, Electron, Python, Go, Java, APIs, data systems, distributed infrastructure, NeMo, NIM, Nemotron, TensorRT-LLM.
Responsibilities
Improve reliability, performance, and observability of AI products; design and build scalable distributed services; establish patterns for CI/CD, safe rollout, and operational excellence; build reusable agent capabilities; codify architecture and mentor engineers; define core architecture for agent collaboration and governance; partner with domain teams to deliver measurable outcomes.
Seniority
Principal, hands-on engineering leader with architectural vision and mentorship.