Skip to main content
Question

Clarification on using Strava activity data for personalized training analysis

  • September 18, 2026
  • 1 reply
  • 6 views

Forum|alt.badge.img

Hello Strava Developer team,

I am developing CycloCoach AI, a personal cycling training and analysis application currently preparing for a very small private beta.

The application would allow an athlete to connect their own Strava account through OAuth and retrieve their own activities in order to:

  • display their training history;

  • calculate personal metrics such as training load, progression, records and consistency;

  • compare planned sessions with completed activities;

  • provide personalized explanations and training recommendations;

  • adapt the athlete’s future training plan based only on their own activity data.

The data would remain private to the connected athlete. It would not be sold, publicly shared, used for advertising, or made visible to other users.

We would also not use Strava data to train a general-purpose AI or machine-learning model.

Before opening the private beta, we would like clarification on the following points:

  1. Is the use of Strava API activity data for individual performance analysis and personalized training-plan adjustments permitted?

  2. May derived metrics such as training load, progress indicators, personal records and training recommendations be stored for the athlete?

  3. If an external AI service is used only to generate a personalized explanation or recommendation, is this permitted when:

    • only the minimum necessary information is transmitted;

    • the information is used exclusively for that athlete’s request;

    • it is not used to train the external provider’s models;

    • no Strava access or refresh token is transmitted?

  4. What retention rules should be applied to raw Strava activity data and to derived training metrics?

  5. Would separate, explicit athlete consent allow anonymized and aggregated statistics to be used later for improving the application, or must all Strava-derived information remain excluded from this type of analysis?

Our goal is to ensure full compliance before allowing any external tester to connect their Strava account.

Project information: https://augmenta-ai.fr/#about

Thank you for your guidance.

1 reply

Jan_Mantau
Forum|alt.badge.img+26
  • Hub Powerhouse
  • September 19, 2026

Some of your questions are answered in the Strava API policy on https://www.strava.com/legal/api_policy

In general you can assume that using AI in any way will lead to a rejection in the review. Exceptions are unknown.