Fittir - Running Shoe Assistant
Web Development

Fittir - Running Shoe Assistant

A modern, full-stack web-application helping runners find the right running shoe online.

Completed: October 15, 2024

Key features

Programmatic SEOAI AgentsRecommendation EngineProduct Stock IntegrationsMobile OptimisedWeb scrapers

Fittir - A platform helping runners find the right running shoe online.

Fittir helps runners find the right running shoe online, helping them search, compare and get recommended across all brands and models based on their needs and goals.

Overview

Fittir is a modern, full-stack web application supplemented by a backend server that handles complex and intensive daily processing activities.

In the front-end, Fittir is has an easy to use questionnaire, advanced filtering and comparison tools, a complex recommendation engine and review system.

In the backend, 1000's of shoes are upkept with standardised data, while a number of AI agents maintain a database of 10,000's of partnered deals (stock integrations) and a dynamic database of reviews about the shoe from youtube, google and Fittir. Online reviews are processed by AI to extract trends and key shoe details to then make recommendations about more than just the paper stats.

Since launching and bootstrapping since 2022, Fittir has served over 10,000 runners with recommendations, and is now a profitable business commercialised by affiliates, widgets for retailers and data services.

Key Features

  • Recommendation Engine: A highly detailed shoe recommendation engine.
  • Payment Integration: Secure payment processing with Stripe
  • AI Agents: A team of 5+ AI agents manages reviews, product data and retailer integrations.
  • User Authentication: Secure user registration and login system to save profiles.
  • Comparison Page: Detailed page for comparing shoes side by side.
  • Challenges Solved

    One of the main challenges was implementing a real-time inventory management system that could handle concurrent orders while maintaining data consistency. I solved this by implementing optimistic locking and database transactions.

    Results

    The platform achieved:

  • • 95+ Lighthouse performance score
  • • 40% faster load times compared to the previous solution
  • • 99.9% uptime since deployment
  • • Positive user feedback on the checkout experience