Lead Evaluation Flywheel Engineer – Autonomous Driving

Nvidia Corporation; Santa Clara (CA); On-site; USD 224,000 - 356,500; Full time

Benefits offered by this job

Equity

Benefits

Job summary

NVIDIA in Santa Clara is seeking a senior software leader to own the eval flywheel strategy and architecture, guiding how driving data becomes curated golden datasets and how metrics are measured against them for trust. You will build self-serve pipelines and reporting tools, and partner with senior engineers across test engineering, behavior planning, and infrastructure to accelerate evaluation cycles.

Collaborate with model developers to push evaluation forward using learned, VLM-based

Qualifications

• Track record of independent execution and technical leadership.
• Clear, proactive communication with engineers and senior leaders at a fast pace.
• BS or MS in Computer Science, Robotics, or related field.
• 12+ years building software in autonomous vehicles, robotics, or large-scale ML systems.
• Deep experience evaluating ML or robotic systems: metric design, ground-truth/dataset curation, precision/recall.
• Strong Python and data engineering skills for production-scale pipelines.

Responsibilities

• Own the eval flywheel strategy and architecture for curated datasets and metrics.
• Set standards for evaluation quality, dataset health, and release processes.
• Build tooling for self-serve dataset pipelines and metric reporting used daily.
• Partner with senior engineers across test engineering, behavior planning, and infra.
• Collaborate with AI model developers to speed evaluation iterations.

Skills

Independent leadership

Technical leadership

Python

Data engineering

Education

BS/MS in Computer Science, Robotics, or related field

Job description

NVIDIA in Santa Clara is seeking a senior software leader to own the eval flywheel strategy and architecture, guiding how driving data becomes curated golden datasets and how metrics are measured against them for trust. You will build self-serve pipelines and reporting tools, and partner with senior engineers across test engineering, behavior planning, and infrastructure to accelerate evaluation cycles.

Collaborate with model developers to push evaluation forward using learned, VLM-based