Also published at https://repsens.ai/subprocessors.
| Subprocessor | Purpose | Data processed | Location |
|---|---|---|---|
| Anthropic, PBC | AI inference — coaching text, daily briefs, training marking and drill grading, mapping suggestions | Performance metrics, practice transcripts and skill-exercise responses, first names or identifiers as prompt context | United States |
| Supabase, Inc. | Managed PostgreSQL database and authentication | All stored Customer Data | United States |
| Render Services, Inc. | Application hosting and compute | Data in transit and in process | United States |
| Wildbit / ActiveCampaign (Postmark) | Inbound email intake of CRM/DMS reports; operational alerts | Report attachments and message metadata | United States |
Conditional — voice features only. Where an Authorized User enables voice input, their browser vendor (for example, Google LLC for Chrome-based browsers, Apple Inc. for Safari) may process spoken audio to perform speech recognition on the user's device or its own infrastructure. This processing occurs between the user and their browser vendor under that vendor's terms; Provider does not receive the audio. Provider identifies this here for transparency rather than as an engagement of a Subprocessor by Provider.
Model training. Provider has contracted for terms under which Customer Data submitted to its AI Subprocessor for inference is not used to train that Subprocessor's models.
These are the service providers that process data on our behalf. Full terms: Data Processing Addendum.