Product Manager

Microsoft Microsoft · Big Tech · Redmond, WA +1 · Product Management

Product Manager for Windows 365 for Agents, a platform for AI agents to autonomously execute computer-use tasks within a secure, cloud-hosted Windows environment. The role involves defining APIs, runtime primitives, safety guardrails, and admin controls for agentic computing at enterprise scale, working with frontier models and agent frameworks to deliver shipping product.

What you'd actually do

  1. Define and ship features that let AI agents reliably observe screen state, invoke applications, handle authentication, persist context, and complete multi-step tasks inside Windows 365, while giving IT admins the controls they need to deploy agents safely.
  2. Go deep with prompt evaluation, model behavior analysis, telemetry-driven product iteration, and benchmark design.
  3. Work directly with customers, ISVs, and agent developers to gather requirements, identify failure modes in real workloads, and prioritize the platform investments that unlock new agentic scenarios.
  4. Collaborate closely with the Windows 365 engineering team and partner across the company — Copilot and Microsoft AI, Azure AI Foundry, Windows, Microsoft 365, Microsoft Intune, Entra, Security & Responsible AI, and Azure infrastructure.
  5. Work as a feature PM to define the vision and value proposition for agent-facing capabilities, write product specs grounded in agent evaluation data, drive feature crews to deliver and measure outcomes, and validate solutions against real agent workloads.

Skills

Required

  • Product or program management experience shipping software products
  • Hands-on experience building, integrating, or evaluating AI / LLM-powered features
  • Experience with one or more of: operating systems and virtualization, identity and security, enterprise SaaS, developer platforms / SDKs, or AI evaluation and red-teaming

Nice to have

  • Hands-on experience with large language models (LLM), agent frameworks, or computer-use / browser-use agents (e.g. building prompts, tools, evals, or end-to-end agent workflows)
  • Ability to write code or scripts (Python, TypeScript, or similar) well enough to prototype, run evaluations, and reason about API design
  • Strong product intuition for agentic experiences — designing for non-deterministic systems, defining success metrics, and shipping safely in the face of model behavior change
  • Able to turn fast-moving AI research into crisp specs, customer-facing narratives, and decision documents for executive review
  • Comfort with ambiguity, bias for action, and a builder’s mindset

What the JD emphasized

  • hands-on experience building, integrating, or evaluating AI / LLM-powered features
  • Hands-on experience with large language models (LLM), agent frameworks, or computer-use / browser-use agents (e.g. building prompts, tools, evals, or end-to-end agent workflows)
  • Ability to write code or scripts (Python, TypeScript, or similar) well enough to prototype, run evaluations, and reason about API design — this is a builder role, not a purely strategic PM role
  • Product intuition for agentic experiences — designing for non-deterministic systems, defining success metrics, and shipping safely in the face of model behavior change

Other signals

  • AI agents
  • operating systems
  • cloud infrastructure
  • enterprise scale
  • shipping product