CalFit — Personal Fitness & Nutrition Coach

Built by AOT AYO — Full-stack fitness & nutrition coaching platform with AI integration.

A React Native (Expo) mobile app for calorie tracking, workouts, sleep monitoring, intermittent fasting, AI-powered coaching & journaling, and accountability partners.

---

Features

Feature	Description
Calorie Tracking	Log meals with macros, search Open Food Facts + Nigerian food DB, track water intake
Workout & Activity	Exercise catalog, timer/stopwatch with TTS, personal records, quick-start routines
AI Coach	Generate personalized workouts via NVIDIA Llama 3.1; conversational chat for health advice, meal plans, form tips
Food Scanner	AI vision-based food recognition using camera — auto-detect food items with nutrition info and log to meals
Step Tracking	Pedometer integration with live step count, daily goals, calorie conversion
Sleep Tracking	Log sleep hours with quality rating, 7-day history
Intermittent Fasting	Protocol selection (16:8, 18:6, 20:4, 24hr, 5:2), active timer with progress ring
Streaks	Daily check-in streaks with milestone celebrations (3–100 days)
Progress Dashboard	Period-based charts for calories, workouts, water, sleep, steps, body measurements
Accountability Partners	Add up to 3 partners by CalFit ID, compare streaks, set shared goals, real-time chat
Notes	CRUD journal for personal notes, synced to Supabase
AI Coach Chat	Conversational interface with the AI coach — ask health questions, generate meal plans, workouts, get motivation
Notifications	In-app notification inbox with filtering by type
Subscription	Free / Pro / Premium tiers (IAP stubbed for future store deployment)
Data Export	Download all user data as CSV or text report

---

Tech Stack

Layer	Technology
Mobile	React Native 0.81 + Expo 54
Language	TypeScript 5.9
Navigation	React Navigation v7 (Stack + Bottom Tabs + Radial Menu)
State	Zustand v5
Backend/Auth	Supabase (Auth, PostgreSQL, Realtime, Storage)
AI	NVIDIA Llama 3.1 70B (chat) + Llama 3.2 90B Vision (food scanner)
Food DB	Open Food Facts API + local Nigerian food database
Voice	Deepgram STT (speech-to-text for AI Coach)
Animations	React Native Animated + Lottie
Charts	Pure React Native (no external chart library)

---

Project Structure

calfit/
├── App.tsx                       # Root — fonts, auth init
├── index.ts                      # Expo entry
├── src/
│   ├── navigation/               # AppNavigator — AuthStack + AppStack + TabNavigator
│   ├── screens/                  # 15+ screen groups (onboarding, dashboard, calorie, etc.)
│   ├── components/               # Reusable UI (ChatBubble, RadialMenu, WorkoutForm, etc.)
│   ├── modules/
│   │   ├── accountability/       # Partner management + real-time chat
│   │   └── shared/               # AndroidSafeView, UserAvatar, EmptyState, ResponsiveScreens
│   ├── store/                    # Zustand stores (auth, theme, aiCoach, mealPlan)
│   ├── services/                 # Supabase, NVIDIA, Deepgram, food search, etc.
│   ├── hooks/                    # useSteps, useWorkoutVoice
│   ├── theme/                    # Design tokens (colors, typography, spacing)
│   ├── types/                    # TypeScript definitions
│   └── utils/                    # AI prompts, JSON parser, strip-markdown
├── assets/                       # App icons & splash screen
├── supabase/                     # Database migrations
├── app.json                      # Expo configuration
├── eas.json                      # EAS Build profiles
└── package.json

---

Try It Out

Web

View the APP live on your browser with no authentication needed: CalFit Web Version

Android (APK)

Download the latest APK: Google Drive link (link will be updated after build)

iOS (Expo Go)

1. Install Expo Go from the App Store
2. Scan the QR code below with your iPhone camera:

[QR code image will be displayed when expo start is running]

Or clone and run locally:

git clone https://github.com/adiiaot/calfit.git
cd calfit
npm install
npx expo start

---

Running Locally

npm install
npx expo start

Scan the QR code with Expo Go (iOS) or the Expo app (Android).

Note: The Recap screen's image export feature requires the native react-native-view-shot module, which is only available in the Android APK build. On iOS (Expo Go), it falls back to text sharing.

---

Key Architecture Decisions

• Steps tracking lives in TabNavigator (not HomeScreen) so the pedometer subscription survives navigation — all screens read liveSteps from Zustand
• Onboarding flow is owned entirely by OnboardingScreen — authStore.setSession never touches isOnboarding
• AI cache — workout generation results are cached for 5 minutes to avoid redundant API calls for identical params
• Notifications — all in-app notification types are defined in notificationService.tsx and stored in Supabase
• Responsive design — uses ResponsiveScreens.ts with scaleX, scaleY, moderateScale for adaptive layouts across device sizes
• Safe area — custom AndroidSafeView wrapper handles status bar, notch, and home indicator on both platforms

---

Database Schema (Supabase)

Table	Purpose
profiles	User profile, goals, body stats, preferences
food_logs	Daily meal entries with macros
water_logs	Water intake records
workout_sessions	Completed workout sessions
step_logs	Daily step counts
sleep_logs	Sleep duration and quality
fasting_logs	Intermittent fasting sessions
notifications	In-app notification inbox
ai_generated_workouts	Saved AI-generated workouts
ai_api_usage	AI API usage logging
partner_messages	Real-time accountability chat messages
partners	Accountability partner relationships
notes	User journal entries

---

License

This project is open-source for personal and educational use only.

Commercial use (rebranding, white-label, selling to clients, app store publication) requires a paid Commercial License. See LICENSE.md for full terms.

Use Case	Allowed
Viewing/forking for learning	✅
Personal projects	✅
Showcasing in portfolio	✅
Rebranding & selling to a client	❌ Requires license
Publishing on App Store / Play Store	❌ Requires license
Using as a foundation for a commercial product	❌ Requires license

Purchase a Commercial License:

• Buy Full Source Code Setup on Selar: CalFit Licensed Source Code
• Custom / Enterprise: aotnetworklabs@gmail.com

---

Portfolio & Verification

This app was built by AOT AYO UNDER THE AOT Network. Source code is available on GitHub for verification:

• Portfolio: aotnetwork.vercel.app

---

Support

Love the product? Buy me a tip to support ongoing development.

Buy me a coffee ☕