Building a Personal Training Dashboard with the Strava API

If you can't measure it, you can't improve it. Logging every run on Strava is a solid habit, but the real value comes from seeing all that data together over time. That's why I connected my Strava account to Claude via the API and built myself a personal training dashboard.

Where the Idea Came From

This project is the fitness side of something I've been calling "Baran's Second Brain." I originally built the Second Brain as plain terminal output; over time I wanted to turn that raw text into a visual panel. You can build the same integration independently from scratch — you don't need the broader Second Brain setup to follow along.

What the Dashboard Shows

The panel is divided into several sections:

• Running volume: Weekly, monthly, and yearly total distance displayed in three separate cards.
• Longest run: Highlighted as a motivational reference point.
• Days since last run: An at-a-glance indicator of how long you've been inactive.
• Group activity: Of the last 15 events in your running group, how many did you actually attend? I made it to only 5 of 15.
• Weekly volume chart: Shows whether your current week is above or below your all-time account average.
• Next run suggestion: The AI looks at your recent history and recommends a gradual increase — for example, after a long break, it advised me not to jump more than 10% above my previous distance.
• Monthly volume trend: I can clearly see that I run much less in winter and pick up significantly in summer.
• Aerobic efficiency: Measures how many meters your heart covers per beat. This was a metric I hadn't tracked before.
• Personal records and weekly rhythms: Useful for understanding where you stand relative to your own history.
• Recent activities: A summary list of your latest runs.

The visual style is Neo Brutalism, which I simply described to Claude and it applied.

How to Set It Up

The integration is straightforward:

1. Log in to Strava, then go to Profile → Settings → My API Application.
2. Create a new API application; Strava will give you a Client ID and Client Secret.
3. Pass these credentials to Claude (or whichever AI tool you're using) and connect the accounts using the access token.
4. All your activity data from the day your account was created will be pulled in automatically.
5. Then prompt the AI: "Build me a dashboard in a Neo Brutalism style showing weekly distance, monthly volume, personal records…" The content and layout are entirely up to you.

Although this walkthrough uses running as the example, the same approach works for cycling, swimming, or any other sport Strava supports.

What Comes Next

Moving from terminal output to a visual dashboard was a meaningful step. The next goal is turning this into a voice assistant — giving commands verbally, hearing the data read back, and eventually running it on a small physical device sitting on a desk. That's still in progress, but the direction is clear.

Seeing your data in one place makes it easy to notice patterns: when you're most active, how long your breaks actually are, and how you've changed over time. That awareness is the first step toward training smarter.