Stride Sync
A workout tracking platform with an AI coach designed to help runners plan, track, and analyze their training with personalized plans and detailed analytics.
Project Details
Client
Personal Project
Year
2025
Role
Full Stack Developer
Technologies
The Challenge
As I prepared for the New York City Marathon, I discovered that existing running apps lacked flexibility for customizing marathon training plans to fit individual schedules and preferences. Most apps offered rigid, pre-made plans with little room for personalization. Additionally, I wanted to challenge myself technically by integrating AI capabilities to create an intelligent training assistant that could generate personalized plans based on a runner's specific needs.
The Solution
I built StrideSync as both a practical solution for my marathon training and a technical showcase. The application combines a flexible workout calendar with a Claude-powered AI coach that generates customized training plans as structured data. The system translates natural language conversations about fitness goals into actionable workout schedules, offering the personalization missing from commercial running apps while demonstrating advanced integration between modern web technologies and AI capabilities.
Technical Implementation
Backend Infrastructure
- •Supabase for authentication, database, and row-level security policies
- •Next.js App Router with server components, client components, and API routes
- •Cookie-based authentication with SSR support
- •PostgreSQL database with specialized workout subtypes
Frontend Framework
- •React 19 with functional components and hooks
- •TypeScript for type safety and enhanced developer experience
- •shadcn/ui and Tailwind CSS for accessible, customizable components
- •Responsive design with light/dark theme support
Performance Optimizations
- •Dynamic imports for improved page load performance
- •Streaming responses for AI interactions
- •SSR for critical pages and CSR for interactive components
- •Vercel deployment with serverless functions
AI & Data Security
- •Anthropic Claude API integration with structured JSON outputs
- •Custom prompting system for workout plan generation
- •Row-level security policies ensuring user data privacy
- •Flexible data modeling supporting various workout types
Key Features
AI Coach
Personalized training plans generated by Claude AI with intelligent workout recommendations based on user goals and fitness level
Performance Analytics
Interactive data visualizations showing workout trends, mileage tracking, and completion rate metrics
Smart Calendar
Interactive workout scheduler supporting different workout types with comprehensive tracking capabilities
Lessons Learned
This project deepened my understanding of technical concepts and implementation strategies. Key insights include:
- 1.Effective prompt engineering is crucial for creating AI systems that generate structured, actionable outputs
- 2.A well-designed database schema with specialized subtypes allows flexibility while maintaining query efficiency
- 3.Balancing server-side and client-side rendering based on component needs significantly improves user experience
- 4.Row-level security implementation is essential for building multi-user applications with proper data isolation
Project Gallery
