(usa) Senior, Data Scientist - Digital Twin Systems Developer

Walmart · Retail · Bentonville, AR

Senior Data Scientist to develop a digital twin platform for Walmart's retail operations. This role involves building modular simulation engines using AnyLogic and Java, creating 3D pipelines for real-time environments, and integrating agentic frameworks driven by LLM-based reasoning or RL. The goal is to enable autonomous, data-driven decision-making by syncing live store data into digital twin modules.

What you'd actually do

  1. Lead the development of AnyLogic models using a library-based approach, creating interchangeable process blocks (via XML and Java) that represent discrete retail pillars like "Last Mile," "Backroom Ops," and "Front-End Checkout."
  2. Create and manage a pipeline for converting and optimizing 3D CAD assets into real-time ready formats (USD, FBX) for deployment in NVIDIA Omniverse, Unreal Engine, or Unity.
  3. Design and deploy "Agentic" frameworks where simulation entities (associates, customers) are driven by LLM-based reasoning or Reinforcement Learning (RL) rather than static scripts.
  4. Standardize simulation inputs/outputs by developing custom Java classes and XML schemas to ensure model scalability and cross-functional portability.
  5. Collaborate with IoT and Telemetry teams to feed live store data into digital twin modules, enabling "shadow" simulations that mirror current physical state.

Skills

Required

  • AnyLogic Professional
  • Java
  • Multi-Agent Systems (MAS)
  • Agentic frameworks (e.g., LangChain, AutoGPT, or RL-based agents)
  • 3D modeling tools (Blender, Maya, or CAD software)
  • XML/JSON structures
  • Modular programming
  • Version control for 3D assets

Nice to have

  • Python
  • Spark
  • Scala
  • R
  • scikit learn
  • tensorflow
  • torch
  • Machine learning
  • Optimization models
  • NVIDIA Omniverse
  • Unreal Engine
  • Unity
  • LLM-based reasoning
  • Reinforcement Learning (RL)
  • Universal Scene Description (USD)

What the JD emphasized

  • AnyLogic Professional
  • Java
  • Multi-Agent Systems (MAS)
  • Agentic frameworks
  • LangChain
  • AutoGPT
  • RL-based agents
  • Universal Scene Description (USD)

Other signals

  • agentic workflows
  • LLM-based reasoning
  • Reinforcement Learning (RL)