HealthIQ AI
Scan your food, know your health — an AI nutrition coach in your pocket.
People want to eat better but have no fast, trustworthy way to judge what is actually on their plate. Reading nutrition labels is slow and meaningless without context: "9 g of sugar" says nothing about whether that food fits a diabetic on a weight-loss goal or an athlete bulking. Traditional calorie trackers make the user do all the work — search a database, guess a portion, log a row — and still return raw numbers with no judgement, no personalisation, and no allergen safety. Most people abandon them within a week, and the ones built for Western supermarkets miss regional foods entirely.
HealthIQ AI turns a phone camera into a personal nutrition analyst. Point it at a meal, a packet, or a barcode and Claude returns a full breakdown in seconds: identified food, portion-aware macros and micronutrients, a personalised 0–10 health score, per-nutrient flags, allergen warnings, and healthier swaps — all scored against that specific user's goal, medical conditions, and allergies rather than a generic RDA. Around the scanner sits a complete health system: a diary, an AI coach that answers questions with the user's own logged data in context, AI-generated diet plans, fasting and exercise tracking, streaks, and friends. A cost-aware hybrid engine routes known packaged foods through a deterministic scorer and free public food databases, so the expensive AI path is spent only where it adds real judgement.
Outcome
Shipped as a production cross-platform app on a live Firebase backend — 20+ Cloud Functions deployed, server-enforced entitlements and usage caps, a hosted internal admin console with live business metrics, and a Claude-powered analysis pipeline optimised from ~20 s to ~13 s per scan. The product is feature-complete through V2.4 with a subscription tier (US$5.99/mo, $49.99/yr) wired end-to-end through RevenueCat and validated against real billing events in sandbox. It is currently in store-release: Android in internal testing and iOS working through App Review, with early accounts on the platform and monetisation not yet switched on for real revenue.
Key features
- AI food scanning — photo, barcode, or nutrition label, analysed in seconds
- Personalised 0–10 health score computed against the user's goal, medical conditions, and allergies
- Safety-critical allergen detection that caps the score and warns before logging
- Full macro and micronutrient breakdown with per-nutrient ok/moderate/high flags
- Healthier alternative suggestions and a plain-language AI verdict for every scan
- Free unlimited food search over Open Food Facts and USDA FoodData Central with no AI cost
- Deterministic no-AI scoring path for foods whose nutrition is already known
- Editable scan results — correct the name, serving, or nutrients and the change persists
- Daily food diary with quick-log, copy-meal, custom foods, and saved recipes
- AI health coach that answers questions using the user's own logged history
- AI-generated personalised diet plans with weight-progress feedback
- Intermittent fasting tracker with reminders
- Manual exercise logging and workout tracking
- Apple Health and Health Connect sync
- Streaks, gamification, and friends-lite social cheering
- Daily insights, weekly reports, and push meal/fast reminders
- Subscription paywall with 7-day trial, server-enforced premium entitlements and fair-use caps
- Email, Google, and Apple sign-in with full account deletion and data erasure
- Internal web admin console with live user, usage, and revenue metrics and role-based access
Tech stack
Build something similar?
Tell us about your project. We'll share how we'd approach it.
Start a project