Hi Strava dev team / community,
I'm in early planning for a Developer Application and want to confirm a couple of compliance points before I invest further build time. I already opened a support ticket (#16518) but got an automated response redirecting here, so posting the specific questions publicly in case a staff member or someone with experience can weigh in.
The concept: a Developer Application that reads a single authenticated athlete's own Strava data (OAuth, read-only, displayed only back to that same athlete) and uses it to generate a physical, printed product mailed to them — e.g. a periodic printed card or an annual summary. No live dashboard, no data shown to or about other users, no intermediary platform.
Two specific questions:
1. Given the Agreement's language that Developer Applications should not compete with or replicate Strava functionality — does a product centered on a physical/mailed artifact (rather than an in-app digital experience) fall within what's permitted, or would this be considered too close to existing Strava features (e.g. Year in Sport)?
2. If part of the pipeline uses an LLM at inference time to turn an athlete's own derived stats into natural-language captions for their own printed product (no training, no fine-tuning, no data shared with other users) — does that fall under the Section 5.3 AI restrictions, or is that section aimed specifically at training/RAG/embeddings use cases?
Happy to share more detail if useful. Appreciate any guidance from staff or anyone who's gotten clarity on similar questions.
Thanks!