Principal AI Researcher

Workday Workday · Enterprise · Pleasanton, CA +2

Workday AI Research is a newly built, elite organization dedicated to advancing the state-of-the-art in AI to power the Agentic Enterprise. As a Principal AI Researcher, you will be a technical founder and visionary, defining research roadmaps for agentic reasoning, long-horizon planning, and model alignment. This role bridges the gap between groundbreaking theory and massive-scale application, with a focus on foundational model research (pre-training, post-training, alignment) and leading contributions to the AI community through publications and open-source releases. You will also partner with engineering to integrate research into production-ready services and mentor the research team.

What you'd actually do

  1. Define research direction in Agentic AI and LLMs.
  2. Advance the state of the art in agentic systems, including retrieval, grounding, memory, context, personalization, etc.
  3. Foundational Model Research: Advance Workday’s proprietary capabilities in pre-training, post-training (RLHF, DPO), and domain-specific alignment for HR and Finance workflows.
  4. Publish & Open Source: Lead Workday’s contribution to the global AI community through publications at top-tier venues (NeurIPS, ICLR, ICML, ACL, etc.) and strategic open-source releases.
  5. Research-to-Production Bridge: Partner with Agent Factory engineering pods to integrate SOTA research into production-ready services that scale to the world's largest organizations.

Skills

Required

  • AI/ML research
  • deep learning frameworks (PyTorch, JAX, or TensorFlow)
  • LLMs or VLMs
  • training
  • alignment (RLHF/DPO)
  • agentic reasoning frameworks
  • reinforcement learning
  • tool-use
  • multi-agent systems
  • Systems Thinking
  • Leadership in Ambiguity
  • Communication

Nice to have

  • founding member of this research arm
  • publication strategy
  • fundamental intelligence layer of the Workday platform
  • long-horizon planning
  • model alignment
  • responsible, intelligent enterprise agents
  • retrieval
  • grounding
  • memory
  • context
  • personalization
  • domain-specific alignment for HR and Finance workflows
  • strategic open-source releases
  • enterprise data privacy
  • explainable AI
  • fair AI
  • safe AI

What the JD emphasized

  • proven history of taking products from applied research to global-scale production
  • Ph.D. or Master’s degree in Computer Science, Artificial Intelligence, Math, Physics, or equivalent technical field
  • deep expertise in LLMs or VLMs, specifically in training, alignment (RLHF/DPO), or agentic reasoning frameworks
  • SOTA Publication Record: A strong history of lead-researcher publications in top AI venues OR a proven track record of shipping industry-defining AI innovations to millions of users.

Other signals

  • advancing the state-of-the-art in AI
  • fundamental breakthroughs
  • research-to-reality team
  • founding member of this research arm
  • technical founder and visionary