KnowEat

Context

Eating out with allergies or intolerances is quietly stressful. Menus are inconsistent, staff can't always answer, and a small mistake carries real consequences.

Challenge

Turn a physical menu into information a person can trust — instantly, privately, and without asking them to read every ingredient list before ordering.

My role

Product concept, interaction design, SwiftUI engineering, App Store release.

Process

I moved between paper sketches, SwiftUI prototypes, and real-world testing in restaurants. Each iteration removed a step, a screen, or a decision.

Design exploration

I explored several visual grammars for scan results — icons only, color-coded lists, layered cards — and landed on a calm, hierarchical layout that surfaces what matters most.

Technical execution

Built end-to-end in SwiftUI with Vision-powered OCR. State is scoped tightly to each scanning session. Preferences persist locally, never leave the device.

Final solution

A single tap opens the camera; a scan returns a clear compatibility view; users can drill into a dish for detail or move on with confidence.

Outcome

Shipped to the App Store as a solo product — taken from initial idea to a live, downloadable app.

Learnings

Constraint is a design tool. The more I removed, the more the product felt like a companion instead of an assistant.

Key product decisions

• Make the scanner reachable in one action from the first screen.
• Use visual indicators — not paragraphs of text — to communicate compatibility.
• Treat allergies, intolerances, and dietary restrictions as first-class citizens.
• Avoid collecting personal data; keep everything on-device where possible.
• Treat accessibility as a product requirement, not an afterthought.