Staff Product Manager, Generative AI Platform

London · London, United Kingdom - Deliveroo · 8+ years experience · Full-time · Posted Oct 20, 2026

At a glance

We're looking for a Staff Product Manager in London to own major parts of that platform.
Your customers are the engineers and ML researchers across all three brands who build AI products and agents.
You'll work with engineering and product leaders in Europe and the US, and your decisions will shape how AI gets built across every market we operate in.

Set the strategy and roadmap for core parts of our GenAI platform: model access and routing, agent infrastructure, and evaluation and observability.
Build one platform for three brands by spotting patterns across Deliveroo, Wolt and DoorDash teams and turning them into shared building blocks.
Shape our model portfolio across frontier and open-weight models, deciding where to build, buy or partner.
Make agents production-ready by giving teams safe access to company knowledge, tools and systems with quality, latency and cost controls.
Own the evaluation strategy so teams can prove quality before launch and keep improving it after.
Stay close to your customers by working with teams shipping AI experiences, measuring adoption and time-to-production, and letting real production use drive the roadmap.
Lead at a senior level, presenting to senior technical leadership and influencing strategy beyond your own roadmap.

Significant product management experience (typically 8+ years, or equivalent), including owning a technical platform, developer product, or ML or data infrastructure.
Hands-on experience with LLM-based products or platforms in production, such as assistants, agents, evaluation or inference.
A platform mindset to notice problems five teams keep solving and turn them into one capability everyone can use.
The ability to go deep with engineers and ML researchers, then explain trade-offs in plain language to a senior leader.
A track record of leading large, ambiguous areas on your own and aligning stakeholders across companies, countries and time zones.
Comfort with regular overlap with US time zones, including some late-afternoon and evening calls.

Comfort with regular overlap with US time zones, including some late-afternoon and evening calls.