Senior Engineer, Human-vs-AV Driving Analytics - Autonomous Vehicles

NVIDIA•Us, CA

15d

About The Position

Want to join a fun, creative company that is on the cutting edge of outstanding technologies? NVIDIA is developing groundbreaking solutions in some of the most exciting technology areas globally, including Virtual Reality, Artificial Intelligence, Deep Learning and Autonomous Vehicles. NVIDIA's Autonomous Vehicles Eval and Metrics team compares autonomous driving behavior against human driving baselines across hundreds of thousands of scenarios. The space between "where we drive differently than humans" and "why, and does it matter" is where this role lives. We are looking for an expert engineer to build the analytical layer that converts raw divergence data into actionable engineering signals, and serve as the technical lead for this workstream. In this role you will define the analytical foundation for how NVIDIA measures AV-vs-human driving quality. This is a blend of data science and systems engineering on safety-critical data.

Requirements

PhD with 4+ years, MS with 6+ years, or BS (or equivalent experience) with 8+ years of proven experience in Computer Science, Computer Engineering, or a related technical field.
Strong software engineering experience especially in Python.
Deep experience in at least two of: statistical analysis, clustering / unsupervised ML, time-series analysis, regression testing systems, or data pipeline architecture.
Proven record to build tooling that turns raw data into decision- quality signals — you've built dashboards, anomaly detectors, or regression pipelines that engineering teams actually used to make ship/no-ship calls.
Comfort with large-scale driving datasets (petabyte-scale sensor logs, millions of scenario runs).
Experience with Spark, Bazel-based builds, or similar big-data tooling.
Strong statistical intuition: you can reason about correlation vs. causation, coverage gaps, distribution shift, and when a metric delta is real vs. noise.
Technical leadership experience: you've set direction for a workstream, mentored engineers, and influenced decisions across team boundaries.

Nice To Haves

Experience in autonomous driving, ADAS validation, or human-factors evaluation (comparing human vs. automated driving behavior).
Strong C++ skills.
Familiarity with closed-loop simulation environments and the gap between sim metrics and real-world performance.
Background in forecasting or predictive modeling — especially predicting system-level outcomes from component-level metrics.
Experience building regression test curation systems (auto-generating or auto-triaging test cases from production failures).

Responsibilities

Be responsible for the analytical capability for human-comparison evaluation.
Develop systems that correlate metric divergence between autonomous and human driving with driving context, including map, scenario type, agent behavior, and environmental conditions.
Automate pattern isolation: compose and build clustering systems that group similar regressions to surface candidate failure modes, turning noisy metric deltas into actionable signals for the driving stack teams.
Build the human-comparison-to-regression-evaluation pipeline end-to-end: surface regression candidates from analysis, curate them into clean regression tests, and close the loop so that every meaningful divergence from human driving becomes a supervised, reproducible test case.
Invest in on-road KPI forecasting — develop models that predict real-world driving KPIs from closed-loop and coverage metrics, and validate that simulation-based evaluation actually correlates with on-road quality.
Set technical direction for human-comparison analytics as a discipline — define the roadmap, mentor engineers, and influence cross-team evaluation strategy.
Partner with MLE tools teams to ensure your analytical tooling integrates cleanly with existing data pipelines and evaluation infrastructure.