Shared Swift Core
Nutrition models and platform-neutral service boundaries live once in shared Swift.
NextMeal is a camera-first nutrition companion built around private, on-device intelligence. Aim at a meal or grocery item, carefully review the editable estimate, and save only the values you choose. Recognition and nutrition values can be completely wrong.
The shared Swift app now includes the camera workflow, editable nutrition review, local diary, personal goals, seven-day insights, complete localization, and synchronized release tooling.
Nutrition models and platform-neutral service boundaries live once in shared Swift.
Apple Vision and Foundation Models plus Gemini Nano through Android AICore stay behind one small interface.
No cloud model, remote meal analysis, ad network, or analytics SDK is present.
Shared UI, permission copy, App Store, and Google Play metadata cover 11 languages.
Fastlane structure, metadata generation, audits, and screenshot tooling are checked in.
Swift Testing verifies nutrition rules, local history behavior, model payload safety, and platform mappings.
Each milestone stays useful without forcing a remote account or cloud inference path.
Native camera preview, on-device image understanding, editable nutrition result, and explicit local save.
Personal calorie and macro targets, a private diary, and seven-day logged trends.
Smart swaps, recipes, deficiency indicators, and permission-based health-platform integration remain future work.
NextMeal processes meal images and nutrition reasoning on the device when supported. There is no cloud fallback hiding behind the camera button.