(usa) Staff, Data Scientist

Walmart · Retail · Sunnyvale, CA

Staff Data Scientist to architect the semantic layer for autonomous commerce, transforming retail data into an agent-consumable environment. Focus on building Knowledge Graphs, Embeddings, and Metadata layers to power AI agent reasoning, tool-calling, RAG, and multi-step planning.

What you'd actually do

  1. Lead the scientific strategy for building the context and semantic layer. You will define how structured, semi-structured, and unstructured data are unified into Knowledge Graphs, Embeddings, and Metadata layers that power agentic reasoning.
  2. Act as the technical "North Star" for the organization. You will ensure that every data product built is optimized for Tool-Calling, Retrieval-Augmented Generation (RAG), and Multi-Step Planning.
  3. Constantly scout the frontier of AI research. You will evaluate emerging frameworks (e.g., LangGraph, AutoGen, GraphRAG) and state-of-the-art models to determine which will define the Sam's Club agentic stack.
  4. Interface directly with Engineering Managers and Data Scientists. You are the "technical glue" that ensures engineering pipelines deliver the semantic richness and reasoning density required by our model-driven agents.
  5. Move beyond standard data quality. You will create evaluation frameworks that measure Semantic Precision, Retrieval Relevancy, and Reasoning Trajectory Success, ensuring our agents are grounded and hallucination-free.

Skills

Required

  • Data Science
  • Applied ML
  • Semantic Engineering
  • Agentic Ecosystem Expertise
  • Vector Databases
  • Graph Databases
  • LLM Orchestration
  • Ontology Design
  • Knowledge Graph construction
  • Embedding space optimization
  • Python
  • Java
  • Scala
  • Schema choices
  • latency
  • data density

Nice to have

  • Dataflow
  • Reinforcement Learning
  • Contextual Bandits
  • Ranking & Retrieval

What the JD emphasized

  • proven track record of deploying complex, model-driven systems into production
  • Deep, hands-on experience with the agentic stack
  • Expert-level understanding of Ontology Design, Knowledge Graph construction, and Embedding space optimization
  • Fluency in Python, Java, or Scala
  • Experience leading through influence

Other signals

  • semantic layer
  • agentic action
  • knowledge graphs
  • embeddings
  • tool-calling
  • RAG
  • multi-step planning