Principal, Data Scientist -- Agentic AI

Walmart · Retail · Sunnyvale, CA

Principal Data Scientist to shape the future of Generative AI and Agentic AI at Walmart, architecting and scaling next-generation ML and LLM systems. This role involves building agentic AI systems with capabilities like planning, memory, and tool-use, defining LLMOps best practices, and driving applied research in LLMs and multi-modal AI. Responsibilities include operationalizing RAG pipelines, translating business problems into AI solutions, building end-to-end ML systems, championing MLOps, and providing technical leadership and mentorship.

What you'd actually do

  1. Architect and build agentic AI systems with capabilities such as planning, memory, tool-use, and multi-agent collaboration, ensuring seamless integration with Walmart’s digital ecosystem.
  2. Define best practices for LLMOps, including prompt versioning, embedding store management, synthetic data generation, fine-tuning strategies, and continuous evaluation for safety, accuracy, and bias.
  3. Drive applied research in large language models (LLMs) and multi-modal AI, leveraging methods such as LoRA/PEFT fine-tuning, reinforcement learning from human feedback (RLHF/DPO), quantization, and speculative decoding.
  4. Explore and operationalize RAG pipelines, deep research, multimodal fusion, and specialized architectures for language, vision, time series, and beyond.
  5. Translate ambiguous, high-impact business problems into scalable AI/ML solutions with measurable outcomes.

Skills

Required

  • Generative AI
  • LLMs
  • Agentic AI systems
  • LLMOps
  • fine-tuning
  • RAG
  • multi-agent orchestration
  • applied science
  • ML models
  • production systems
  • large-scale ML/AI platforms
  • Python
  • PyTorch
  • TensorFlow
  • Agentic development platforms (Google ADK, LangGraph, CrewAI, AutoGen)
  • technical leadership
  • mentoring
  • architectural decisions

Nice to have

  • Scikit-learn
  • Spark ML
  • XGBoost
  • Java
  • C++
  • SQL
  • R
  • distributed training
  • model compression
  • ONNX interoperability
  • multimodal systems (vision, speech, time series)
  • AI/ML community (publications, open-source contributions, conference talks)

What the JD emphasized

  • deep fusion of cutting-edge applied science and scalable production system delivery
  • architect and scale next-generation ML and LLM systems
  • pushing the boundaries of what large-scale AI can do in production
  • Deep expertise in Generative AI, LLMs, and/or Agentic AI systems
  • Strong applied science background
  • Proven experience architecting large-scale ML/AI platforms
  • strong coding skills
  • independent coding prowess
  • foundational strength
  • drive to lead our next-era AI journey

Other signals

  • architect and scale next-generation ML and LLM systems
  • architect and build agentic AI systems
  • define best practices for LLMOps
  • drive applied research in LLMs and multi-modal AI
  • explore and operationalize RAG pipelines
  • translate ambiguous, high-impact business problems into scalable AI/ML solutions
  • build and guide end-to-end ML systems
  • champion advanced MLOps practices
  • serve as a technical leader and mentor
  • influence cross-functional platform strategy