Principal Product Manager - Microsoft AI and Copilot

Microsoft Microsoft · Big Tech · Tokyo, Tokyo, Japan · Product Management

Principal Product Manager for Microsoft Copilot's Identity & Expression layer, focusing on how AI agents communicate intent, reasoning, and trust. This role involves defining strategy, owning a reference implementation (Mico), translating AI capabilities into enterprise value, establishing trust models, and leading AI evaluation for expressive and agentic experiences, using Japan as a pilot market.

What you'd actually do

  1. Define Copilot’s Identity & Expression strategy across text, voice, and UI-aware surfaces, including how AI agents express reasoning, confidence, uncertainty, and progress.
  2. Own Mico as the reference implementation of Copilot Identity & Expression, ensuring it evolves as a platform capability rather than a standalone feature.
  3. Translate expressive and agentic AI capabilities into clear enterprise value, such as onboarding, workflow guidance, and reduced cognitive load.
  4. Define enterprise trust models for expressive AI, including governance, admin control, safety constraints, and predictable failure modes.
  5. Lead AI evaluation strategy for expressive and agentic experiences, defining quality bars beyond accuracy: trust, tone, appropriateness, and user confidence.

Skills

Required

  • 7+ years of software product management taking from a user need, a prototype from engineering to market.
  • Hands-on experience delivering AI or generative-AI-powered features or products.
  • Experience working on enterprise or business-facing products, including IT, security, or operational constraints.
  • Experience evaluating product quality using both quantitative metrics and qualitative feedback, including cases where release decisions were adjusted or delayed.
  • Proven experience working cross-functionally with engineering, design, research, and business stakeholders.

Nice to have

  • Experience with AI agents, Copilot-style assistants, or conversational interfaces.
  • Enterprise productivity, collaboration, or workflow automation background.
  • Experience working with global teams across Japan and other regions.
  • Business-level or fluent Japanese language proficiency.

What the JD emphasized

  • AI-native mindset
  • AI uncertainty
  • trust, tone, restraint, and evaluation
  • AI or generative-AI-powered features or products
  • evaluating product quality using both quantitative metrics and qualitative feedback
  • release decisions were adjusted or delayed

Other signals

  • AI agents
  • Copilot
  • enterprise value
  • trust models
  • AI evaluation strategy