Principal, Data Scientist - Digital Twin Systems Architect

Walmart · Retail · Bentonville, AR

The Principal Data Scientist will architect and lead the development of Walmart's Retail Digital Twin ecosystem, integrating simulation engines, 3D frameworks, and agentic AI to optimize retail operations. This role involves designing end-to-end systems, selecting and optimizing the tech stack, driving software engineering excellence for production-grade applications, and leading the strategy for agentic AI workflows.

What you'd actually do

  1. Lead the design of a modular, distributed digital twin architecture that balances simulation fidelity with computational efficiency across diverse retail use cases.
  2. Evaluate and deploy the most effective simulation paradigms (Agent-Based, DES, Physics-based) and tools—whether leveraging AnyLogic, NVIDIA Omniverse, or building custom C++/Python-based engines from the ground up.
  3. Build production-grade, scalable applications that integrate real-time IoT data streams, ensuring the digital twin platform is performant and maintainable at an enterprise scale.
  4. Design the orchestration layer for agentic workflows, enabling intelligent entities within the twin to perform complex reasoning and autonomous optimization.
  5. Define the schema and governance for reusable process blocks (XML/JSON) and 3D CAD assets (USD/gLTF) to ensure interoperability across the organization’s pillars.

Skills

Required

  • System Architecture Mastery
  • Advanced Software Engineering (C++, Java, or Python)
  • Agentic AI & Optimization
  • PhD or MS in Computer Science, Robotics, Computational Physics, or Industrial Engineering

Nice to have

  • Simulation paradigms (Agent-Based, DES, Physics-based)
  • AnyLogic
  • NVIDIA Omniverse
  • custom C++/Python-based engines
  • real-time IoT data streams
  • Kubernetes
  • shaders and ray-tracing
  • Unreal
  • Java-based process logic
  • data structures
  • algorithms
  • distributed systems
  • Reinforcement Learning (RL)
  • Multi-Agent Systems (MAS)
  • Large Action Models (LAMs)

What the JD emphasized

  • architect of systems
  • agentic AI
  • Agent-Based
  • DES
  • Physics-based
  • Agentic AI & Optimization
  • Reinforcement Learning (RL)
  • Multi-Agent Systems (MAS)
  • Large Action Models (LAMs)
  • built a proprietary or highly-customized digital twin platform

Other signals

  • digital twin
  • agentic AI
  • simulation engines
  • optimization