Validation Data Scientist, Verification and Validation - Autonomous Vehicles

NVIDIA NVIDIA · Semiconductors · Shanghai, China +2

The Validation Data Scientist will build tooling, perform large-scale analysis, and drive data-driven evaluation of vehicle-level behavior and Operational Design Domain (ODD) coverage during scaled testing for autonomous vehicles. This role involves building and improving evaluation frameworks, data pipelines, and data curation strategies, defining core metrics, and automating scalable workflows using cloud platforms and AI. The goal is to influence product development, technical reviews, and software releases by providing quantitative analyses and clear reporting.

What you'd actually do

  1. In this role, you will lead large‑scale driving behavior and ODD analysis using extensive real‑world and virtual AV driving logs to evaluate safety, comfort, and overall vehicle‑level performance.
  2. Build and improve evaluation frameworks, data pipelines, and data curation strategies to support robust analysis across thousands of test miles every single day.
  3. Define and compute core metrics that quantify AV performance against target ODDs, powering our product development flywheel, technical reviews, and AV software releases.
  4. Build and automate scalable workflows that use cloud platforms, modern data engineering tools, and AI workflows to surface insights, spot regressions, and enable data‑driven decision making.
  5. Working closely with our Software Product, Testing and Development teams, you will turn open-ended safety and performance questions into clear quantitative analyses that influence what we build next.

Skills

Required

  • MS or PhD in Computer Science, Mathematics, Statistics, Electrical/Computer Engineering, or a related quantitative field, or equivalent experience.
  • 5+ years of proven experience in data science, data engineering, or analytics roles working with large‑scale data, ideally in safety‑critical domains.
  • Experience analyzing behavior of autonomous vehicles, ADAS systems, or other safety‑critical cyber‑physical systems.
  • Strong Python skills, including writing production‑quality code and libraries for data processing, analysis, and automation.
  • Hands‑on experience building and operating data pipelines in a production environment with cloud computing platforms.
  • Excellent communication and teamwork skills, with a track record of working across teams and presenting your findings to senior leaders.
  • Ability to compose clear user documentation, technical guides, and executive‑level summaries.

Nice to have

  • Background in statistics including experimental design, hypothesis testing, confidence intervals, and explaining results for non‑experts.
  • Proven track record architecting and scaling data and ML/AI pipelines to process and analyze very large telemetry or log datasets.
  • Experience with GPU‑accelerated and/or distributed computing for large‑scale data processing and model evaluation.
  • Familiarity with simulation‑based validation, vehicle‑level testing, and interpreting fleet test or on‑road validation data.
  • Experience leading technical direction for a data or analytics team, including setting standards for code quality, metrics, and validation methodologies.

What the JD emphasized

  • large-scale analysis
  • data-driven evaluation
  • evaluation frameworks
  • data pipelines
  • data curation strategies
  • core metrics
  • scalable workflows
  • AI workflows
  • quantitative analyses
  • validation results
  • large-scale data
  • safety-critical domains
  • autonomous vehicles
  • ADAS systems
  • safety-critical cyber-physical systems
  • production environment
  • cloud computing platforms
  • senior leaders
  • data and ML/AI pipelines
  • large-scale data processing
  • model evaluation
  • simulation-based validation
  • vehicle-level testing
  • fleet test or on-road validation data
  • technical direction
  • data or analytics team
  • metrics
  • validation methodologies

Other signals

  • large-scale analysis
  • data-driven evaluation
  • AV software releases
  • scalable workflows
  • AI workflows
  • quantitative analyses
  • sophisticated validation results