Portfolio:

Applied Innovation – Growth & Innovation (ES)

Rank:

Associate Director

Sub Portfolio:

Next Frontier

Reports to (Job Title and name): 

ES Growth & Innovation Lead

Role title:

Physical AI Data Engineering Lead

About the Next Frontier:

Applied Innovation's Next Frontier focuses on emerging technologies, including Physical AI, Quantum Computing, and Next-Generation AI amongst others. This initiative translates advanced capabilities into measurable outcomes, helping clients and EY teams improve decision-making, strengthen resilience, and unlock new opportunities through technology-driven solutions.

About the Delivery Team:

EY advances innovation in Physical AI by enabling regional robotics labs and supporting business-driven initiatives. The focus is on practical Physical AI technologies and framework that deliver measurable, enterprise-value level benefits.

Next Frontier’s Physical AI team is led by Youngjun Choi, EY Global Robotics and Physical AI Leader.

Position Summary:

The Data Engineering Lead will work on Physical AI initiatives, and interlock with other data focused workstreams within Innovation. This role leads the development of AI‑ready data pipelines, synthetic data ecosystems, and digital twin models, enabling organizations to safely design, simulate, validate, and deploy AI‑driven physical automation at enterprise scale.

. By embedding strong data governance and Responsible Physical AI guardrails across the entire lifecycle, the role ensures that Physical AI solutions meet safety, ethical, compliance, and resilience standards while helping industries transition from experimentation to production‑grade automation.  

Essential Responsibilities of the Job:

• Architect end‑to‑end AI‑ready data pipelines that generate, curate, and govern high‑quality datasets required to simulate physical AI scenarios at scale, addressing data quality, accessibility, and consistency challenges
• Lead the creation and maintenance of synthetic data generation systems to model diverse operational conditions and edge cases, ensuring robust, risk‑aware training of robots, drones, and smart‑edge devices
• Build and operate digital twin data models using NVIDIA Omniverse to capture real‑world environment dynamics, enabling data‑driven testing, optimization, and de‑risking of physical AI deployments.
• Oversee the integration of simulation training datasets from NVIDIA Isaac frameworks, ensuring data fidelity and completeness for validating AI‑driven robotics in 3D environments.
• Manage ingestion, transformation, and execution of compute‑intensive AI workloads on NVIDIA AI Enterprise, ensuring data throughput, scalability, and security for training and inference.
• Apply Responsible Physical AI principles by embedding data governance controls i.e., safety, ethics, compliance, resilience, across the full data lifecycle
• Support client showcases using demos/proof of concepts. Enable Physical AI Data related thought leadership

Analytical/Decision Making Responsibilities:

• Evaluate when to use real‑world data vs. synthetic data based on scenario variability, risk exposure, model generalization needs, and gaps in ground‑truth observations.
• Decide appropriate data fidelity levels for digital twin simulations to optimize accuracy, simulation speed, and compute resource usage on accelerated NVIDIA infrastructure.
• Assess data completeness and reliability before approving robotic behaviors for pilot deployment, ensuring simulation‑to‑reality transfer readiness.
• Identify data‑driven safety risks e.g., biases, incomplete synthetic scenarios, insufficient operational coverage. And escalate adjustments aligned with Responsible Physical AI guardrails.
• Determine appropriate storage, movement, residency, and processing strategies (edge vs. cloud vs. on‑prem accelerated compute) based on data sensitivity, latency, and compliance requirements.

Knowledge and Skills Requirements:

• Expertise in synthetic data generation, including domain randomization, procedural simulation, and scenario augmentation, to support scalable physical AI training.
• Deep skill in building and managing digital twin datasets within NVIDIA Omniverse, including sensor modeling, telemetry ingestion, and environment representation.
• Strong command of robotics simulation data workflows using NVIDIA Isaac (dataset creation, sim logs, trajectory data, behavioral modeling) for pre‑deployment validation.
• Proficiency in designing AI‑ready data architectures that meet reliability, scalability, governance, and security standards required for enterprise Physical AI systems.
• Knowledge of safety, ethics, and compliance requirements related to Responsible Physical AI, with the ability to enforce data‑driven guardrails and auditability.
• Familiarity with accelerated data processing and model training using NVIDIA AI Enterprise, including dataset distribution, GPU‑optimized pipelines, and high‑volume simulation data handling.
• Knowledge as well as the ability to strengthen/acquire new knowledge and capabilities to keep up with the continuously evolving technology landscape
• Proven ability to manage high-performing teams and engage stakeholders across various cultures and time zones

Education & Experience:

• MS/PhD in Computer Science, Electronics Engineering, Robotics or related fields
• 7+ years of hands‑on experience with data‑driven robotics or simulation systems, including leadership of programs using Omniverse digital twins and Isaac‑based robotics datasets.
• Demonstrated experience operationalizing synthetic data pipelines and AI‑ready data governance frameworks in real‑world automation or AI deployments.
• Track record delivering data‑intensive AI workloads on NVIDIA AI Enterprise or similar accelerated computing environments.

Note:

This job description is intended as a guide to reflect the principal functions of the job.  However, it is not an all-inclusive listing of the required job functions and functions may vary depending on the geographic location of the job and/or the manager. Further, the job description is subject to change at the discretion of management.