Director, System Software Engineering - Metropolis Accelerated and Inferencing Software

Within NVIDIA's Edge AI, Metropolis, and Blueprints (EMB), this team is the execution engine behind NVIDIA’s Vision AI strategy—owning the full lifecycle from model onboarding to production deployment. We transform foundation models into real‑time, GPU‑accelerated video intelligence systems using Deep Stream and VSS. Our focus includes scaling multimodal reasoning and enabling agentic development workflows. We follow through between production data and model improvement.

This work positions NVIDIA as the default platform for Physical AI.

NVIDIA is looking for a proven Director of Systems Engineering who is hands‑on with deep learning and comfortable reading/modeling code, not just running it. You bring strong intuition for modern architectures (e.g., transformers, diffusion, and VLMs); deep experience tuning on NVIDIA GPUs (kernels, memory, and latency/efficiency trade‑offs) / SOCs; and a consistent track record of delivering robust, low‑latency inference  have led teams that turn Accelerated Computing pipelines into reliable, measurable business impact for embedded and Enterprise platforms.

You will work with a cohesive, high‑performing team that’s been built and refined over the past nine years. An individual well‑aligned with industry experts is a great fit for this role!

• Lead, encourage, and develop world‑class engineering and data teams decentralized across Europe, Asia, and the United States.
• Architect and operationalize NVIDIA’s end‑to‑end data inference acceleration strategy, powering inference and continuous performance improvements.
• Drive strategic implementations of TensorRT, VLLM, and other accelerated frameworks for inference solutions for Edge and Enterprise devices:
  Lead Accelerated Computing efforts and solutions for key Metropolis verticals. Set up Proofs of Readiness (PORs) and guide their implementations.
• Collaborate with major Metropolis OEMs and Partners to architect highly accelerated and optimized custom deep‑learning models and inference pipelines for their specific requirements.
• Offer direct customer support, including debugging, technical education, and handling customer inquiries for our Metropolis partner and customers. Responsible for drafting and finalizing SOWs with internal customers and partners.
• Performance Benchmarking:
  Orchestrate efforts to achieve leading performance results on industry benchmarks like MLPerf on various edge and Enterprise devices.
• Technical Leadership & Influence:
  Function as a technical leader for deep learning across multiple teams, giving oversight and building support. Apply customer insights to shape the composition and structure of upcoming SOC/GPU deep‑learning hardware.
• Scaling the Team:
  Strategically hiring to meet new demands while also mentoring and adjusting existing teams to new deep‑learning challenges.
• Represent Nvidia Deep Learning solutions in webinars, conferences, and partner events.

• Bachelor’s and/or Master's in Computer Science/Electrical Engineering or equivalent experience.
• 15+ years of overall experience, with a minimum of 10+ years of significant involvement in machine learning/deep learning research or practical experience, coupled with 7+ years of leadership background.
• Over 10 years of validated industry expertise in the embedded software sector, holding technical leadership positions accountable for delivering outstanding production software within a multidimensional setting.
• Deep knowledge of GPU, CPU, and dedicated deep‑learning architecture fundamentals, and low‑level performance optimizations using heterogeneous computing.
• Hands‑on experience with VLMs, LLMs, or multimodal AI systems applied to perception, data triage, or automated labeling.
• Strong expertise in large‑scale data processing, systems building, or machine learning pipelines.
• Strong communication, careful planning, and technical leadership capabilities.

• PhD in a relevant field such as Spatial Computing & Awareness, Sim‑to‑Real Transfer, Human-to-Physical AI Interaction.
• Deep experience with CV, LLMs, VLMs, GenAI models, and standards.
• Technical thought leadership in production…