THE AI INTERVIEW
The company has prioritised platform investment first, aiming to equip staff across the business with AI tools rather than restricting access to a central team.
“ Our idea is to empower everyone at JET to do the job they need to do, but with AI,” Liz goes on.“ We’ re trying to transform ourselves into an AI-first company.”
Crucially, every initiative is tested before wider rollout. Liz notes:“ What we ask is,‘ What’ s the smallest thing we can do to test a hypothesis and is it worth it?’. We don’ t just want to build something with AI because it’ s cool.”
On the customer-facing side, Just Eat has launched an AI Voice Assistant, shifting away from traditional search-based navigation towards conversational ordering.
“ It’ s a significant shift for us as we move from a traditional search and scroll interface to conversational commerce,” says Liz.“ It’ s built to understand customer intent, not just keywords.”
The assistant is now live in the UK, with international expansion planned throughout 2026.
Just Eat is also extending ordering into WhatsApp, the messaging platform owned by Meta, with launches planned for the Netherlands and Spain in the first quarter of 2026.
Powering AI with a multi-cloud strategy Running AI at scale requires substantial computing power, particularly from GPUs which handle the intensive calculations AI models depend on.
45 %
Just Eat’ s market share in the UK
Just Eat has adopted a multi-cloud strategy, spreading workloads across multiple cloud computing providers rather than relying on a single vendor.
“ We’ re looking at how we’ re being resilient,” reveals Liz.“ We’ re struggling, like everybody else, with access to GPUs, so we’ re looking at alternatives on what we can do there.”
Cost is a growing consideration too:“ AI can get incredibly expensive to run, so we’ re looking not just at our runtime strategy but also how we’ re sourcing models from a cloud strategic perspective.”
Vibe coding and the route to production One of the standout elements of Just Eat’ s AI strategy is its approach to moving projects from prototype to live deployment. aimagazine. com 29