Dionysios Basdanis

Athletes generate a massive amount of data. Every run recorded on Strava contains signals about performance, fatigue, and progress, but most of that stays locked in charts. As part of the EasyFix team, we built Agon to close that gap—turning raw metrics into actionable guidance.

From Data to Insights

Agon connects to your Strava account via secure OAuth and automatically syncs your recent activities. While our dashboard provides a clear look at distance, pace, and duration, the core value lies in the AI analysis. Using Google Gemini, our system interprets training patterns to help athletes:

• Maintain consistent training volume.
• Balance high-intensity efforts with recovery.
• Avoid overtraining by staying aligned with their current workload.

The Full-Stack Architecture

We built Agon as a modern full-stack web application designed for speed and simplicity:

• Frontend: We used React and TypeScript to provide a responsive dashboard for reviewing activities.
• Backend: A Node.js/Express server handles Strava authentication, data retrieval, and secure user sessions.
• Database: We chose PostgreSQL to reliably store training history.
• AI Engine: We integrated the Google Gemini API to turn existing training data into personalized feedback.

Why We Built This

Elite athletes often rely on coaches to interpret their data, but recreational athletes usually don’t have that luxury. Agon is our experiment in making intelligent training insights accessible to anyone. We didn't want to build a complicated analytics platform—we wanted to create something simple that answers the most useful question: "What should I do next?"