AI Tool

The Last Calorie Tracker

The most frictionless food logger, now native on iPhone.

Type
AI Tool
Platforms
iOS
Status
Live
Role
Design + engineering

The story

Why it exists

Food logging fails when recording a meal takes longer than the habit can tolerate.

My role

Independent product designer and engineer. I defined the logging model, designed the native experience, and rebuilt the product in SwiftUI.

Product move

Unify barcode, voice, text, and nutrition-label capture behind one short review flow, then move recognition onto the phone wherever possible.

What shipped

A native iPhone app with four input modes, on-device Apple Intelligence, macro and weight trends, optional HealthKit integration, and no account or analytics.

See the product

Voice logging turns a spoken meal into structured nutrition without a search-first workflow.
Barcode lookup handles packaged food while keeping the rest of the product local-first.
History makes weight, calorie, and macro patterns visible across useful time ranges.
Guided setup translates goals and activity into editable calorie and macro targets.

Product judgment

Key decisions

  1. 01

    Optimize for the moment of logging

    Every input method ends in the same compact confirmation flow instead of becoming a separate feature journey.

  2. 02

    Put intelligence on the device

    Typed, spoken, and photographed foods are interpreted locally. Only barcode and brand lookups leave the phone.

  3. 03

    Make privacy the default

    Food history and goals stay on-device with no account, tracking, or analytics. HealthKit remains explicitly optional.

Build notes

The original Progressive Web App was rebuilt as a native SwiftUI application with Apple Intelligence handling food recognition on-device.

Open Food Facts is used only for barcode and brand lookup. Face ID protection and HealthKit integration are optional system capabilities.

Built withSwift, SwiftUI, Apple Intelligence, HealthKit, Open Food Facts, Claude Code