Zipli: Designing an interface for kitchens to rescue surplus food at scale.

Hands with gloves holding a phone with a form AI-pre filled UI, warehouse in the background.

AI Context

MY ROLE

I designed Zipli from zero to a €20K pilot. An AI-powered food recognition and matching app for busy kitchens, focused on the hardest surplus to save — warm food.

User-led research

The product design started by rescuing food ourselves and observing the created process - before touching the interface.

User research in kitchens.

Mascot in a pose as its analysing AI Image

Research of surplus foodsharing ecosystem hubs and redistribution depots.

PROJECT BACKGROUND

Every day, good food is thrown away while 37 million people across the EU do not have enough to eat. Zipli connects food producers with people who need it, helping surplus food reach communities instead of landfills.

How might recording food surplus be as easy as turning the stove on?

Superfast donation flow.

A cook scanning leftover pizza slices with phone in industrial kitchen.

1. Snap a photo

AI render

Kitchen staff take a photo of their leftover meals.

AI background

A first person view of a kitchen worker with gloves holding a phone with Zipli app in it

2. AI does the rest

The meals get offered to those in need and are recorded in the restaurants statistics.

SOLUTION

Keep staff focused on the kitchen.

Kitchen staff work in fast‑paced environments. While managing surplus food, they carry huge responsibility for recording hygiene and allergen information correctly. The logging flow is designed to lower that burden. It is based around a single‑step process that captures key details from a single photo.

Data-driven food saving

Designed to work at two levels, the platform gives kitchen managers insight into what is being donated or discarded and why, while helping cities understand food surplus across their communities.

Make AI legible at the moment it matters.

Built a recognisable brand and a practical design system for AI-assisted work. Every interaction makes a simple distinction: what Zipli suggests, what the worker confirms, and what needs attention.

Design system that is built for High contrast, visible primary pathing, greasy hands with gloves and short interaction time. It optimises for clarity, use of AI and food safety.

Friendly, AI-first interface.

Designed a human-ai interaction elements that feel friendly and direct your attention where its needed.