Inspiration
I wanted to create a personal health assistant that makes fitness and nutrition accessible and personalized for everyone. Many apps provide generic plans, but few adapt intelligently to dietary preferences, goals, and available equipment. NutriFit aims to fill that gap with AI-driven recommendations.
What it does
NutriFit generates personalized meal plans, workout routines, and shopping lists. It adapts to user goals, dietary restrictions, and available equipment. Users can interact with a conversational AI to adjust plans, ask questions, or explore alternatives. Everything is mobile-friendly, intuitive, and offline-capable.
How we built it
We used Kiro extensively throughout development. Spec-driven development helped define the core architecture, ChatbotEngine, and PlanParser modules. Vibe coding enabled rapid iteration on UI and edge cases. Agent hooks automated repetitive workflows like route and API generation. Steering documents guided code style, UX design, and error handling. Context management allowed Kiro to maintain consistency across the entire codebase.
Challenges we ran into
Integrating multiple LLMs with seamless fallback logic was complex. Ensuring cross-file consistency and robust error handling required careful design. Balancing offline capabilities with dynamic AI features was another challenge, along with building a conversational AI that understood diverse inputs accurately.
Accomplishments that we're proud of
We built a fully functional AI-powered health assistant with modular, maintainable code. The ChatbotEngine can interpret diverse user inputs and generate structured plans. We achieved end-to-end automation of boilerplate workflows and maintained consistent UX and architecture across all features.
What we learned
We learned how to strategically combine spec-driven development for complex features with vibe coding for iteration. Steering documents and context sharing are essential for maintaining consistency in large projects. Automating repetitive workflows saves time and improves code quality.
What's next for NutriFit
We plan to expand AI capabilities to include predictive analytics, meal cost optimization, and enhanced personalization. Additional features include integration with wearable devices, social sharing of progress, and smarter offline plan caching for uninterrupted user experience.
Log in or sign up for Devpost to join the conversation.