Principal Product Manager - AI Builder

Expedia Expedia · Hospitality · London, United Kingdom

Product Manager for AI Builder role focused on developing and shipping multi-agent systems to automate the merchandising experiment lifecycle for Expedia Group's consumer brands. The role involves prototyping agentic flows, defining trust and governance patterns, and driving measurable outcomes like reduced experiment cycle time and increased conversion.

What you'd actually do

  1. Own the product direction for autonomous merchandising optimisation — from agent-assisted experiment setup today to agents that generate, launch, evaluate, and roll out tests continuously within brand-defined success metrics, guardrails, and policies.
  2. Rapidly prototype agentic flows hands-on: orchestration across specialised sub-agents, tool/function calling against internal platform APIs, prompt and context design, agent memory, and evaluation harnesses — proving value and feasibility before committing engineering investment.
  3. Take successful prototypes to production with engineering, data science, and design — including human review gates, guardrail breach detection, automated rollback, audit trails, and attribution so every agent action is measurable and reversible.
  4. Define the trust architecture for agents acting on live experiences: preview/approve/undo flows, escalation triggers, explainability, and override tracking — patterns designed to be adopted by teams across Expedia Group as autonomous optimisation expands to new surfaces.
  5. Drive measurable outcomes: compress experiment cycle time from months to days, multiply experiment throughput per surface, and demonstrate conversion and revenue lift from agent-optimised merchandising — without degrading traveler trust or experience quality.

Skills

Required

  • product management experience
  • AI, ML, or technically complex systems
  • technical depth
  • partnering with engineering and DS/ML teams
  • agent architectures
  • system design
  • data flows
  • prototyping
  • using APIs/SDKs from major LLM providers
  • writing simple code (e.g., Python or TypeScript/JavaScript)
  • stitch together prompts, tools, and agent workflows
  • experimentation and measurement
  • A/B testing at scale
  • success metrics and guardrail definition
  • instrumentation
  • using experiment data to drive iteration
  • communication and collaboration skills
  • make complex technical topics accessible
  • drive alignment across senior cross-functional stakeholders

Nice to have

  • Experience building or shipping agentic AI systems: multi-agent orchestration, tool/function calling, MCP or similar protocols, agent memory, human-in-the-loop review patterns, and agent evaluation.
  • Background in e-commerce merchandising, personalisation, or marketplace optimisation — pricing display, recommendations, ranking, promotional content, or campaign automation.
  • Experience with auto

What the JD emphasized

  • 10+ years of product management experience
  • Proven track record of taking 0→1 product ideas from concept through prototype to launched features
  • Deep experience with experimentation and measurement
  • Experience building or shipping agentic AI systems

Other signals

  • autonomous agents
  • multi-agent systems
  • experimentation automation
  • trust and governance