SparkyFitness

SparkyFitness is a full-stack, open-source fitness tracking application for logging nutrition, workouts, water intake and body measurements. It provides a web PWA and an Android client, optional Garmin integration, and an AI-powered chat assistant (SparkyAI) for natural-language food logging and image-based meal recognition.

Key Features

• Nutrition logging with custom foods, categories and interactive charts.
• Exercise logging and a searchable exercise database.
• Water intake and daily check-ins to track habits.
• Body measurements tracking with progress charts and goal management.
• AI Nutrition Coach (chat-based food logging, image uploads, chat history).
• Web PWA, Android mobile app, optional Garmin microservice, and Docker-based deployment.

Use Cases

• Individuals or families who want a self-hosted alternative to commercial fitness trackers.
• Users who want AI-assisted food logging and image recognition for meal entries.
• Developers and self-hosting enthusiasts deploying a full-stack app using Docker and PostgreSQL.

Limitations and Considerations

• Several features are noted as beta/under heavy development (AI Chatbot, multi-user/family access, Apple Health sync); expect potential breaking changes and migration work during upgrades.
• The project is intended for self-hosting and requires configuring environment variables, database migrations, and optional external API keys for food providers.

SparkyFitness is a comprehensive, privacy-focused option for users who prefer to self-host their fitness data, offering a modern React front end, Node/Express backend and PostgreSQL storage with Docker deployment support.