Sub-processors
Last updated October 10, 2026
AXO Solutions, Inc. uses the service providers below to run AXO.Voice. Each one receives only what it needs for the purpose shown, under a contract that limits it to providing that service to us. Which providers handle your data depends on which modules you use.
TypeSafe AI is listed as a planned optional provider for the Jev experiment. It is not enabled by default. The notice period below applies before a new provider processes customer personal information.
Current sub-processors
| Provider | What it does | What it receives | Used by | Location |
|---|---|---|---|---|
| Supabase, Inc. | Database, authentication and file storage | All customer data, including recordings, transcripts and account details | All | United States |
| Vercel, Inc. | Application hosting and custom-domain TLS | Requests to the application and the data they carry, in transit | All | United States |
| Deepgram, Inc. | Speech-to-text transcription | Call audio and the vocabulary terms configured for a workspace | AI Call Analyzer, AI Voice Agent | United States |
| Amazon Web Services, Inc. (Amazon Bedrock) | Language-model processing: summaries, dispositions, scorecards and voice-agent conversation turns | Call transcripts, conversation turns and the prompts or grading rules configured for a workspace | AI Call Analyzer, AI Voice Agent | United States |
| Anthropic, PBC | Language-model processing, where a request is not routed through Amazon Bedrock, and conversation turns for the voice agent | Call transcripts, conversation turns and configured prompts | AI Call Analyzer, AI Voice Agent | United States |
| TypeSafe AI, Inc. | Planned optional Jev interpretation of bounded voice-agent answers, when enabled for a script | The current caller utterance, current script question, offered answer options and candidate value spans; no call audio | AI Voice Agent (optional; not enabled by default) | United States |
| ElevenLabs Inc. | Text-to-speech for the voice agent, and voice cloning where a customer enables it | The text the agent speaks; for voice cloning, audio samples of a consenting staff member | AI Voice Agent | United States |
| Twilio Inc. | Telephony and SMS for the voice agent; a recording source where a customer connects its own account | Phone numbers, call audio in transit, call metadata and message content | AI Voice Agent, AI Call Analyzer | United States |
| Google LLC (Google Cloud Run) | Hosts the real-time bridge that connects phone calls to the voice agent | Live call audio and conversation turns, in transit | AI Voice Agent | United States |
| Resend, Inc. | Transactional email: invitations, password resets, alerts and report delivery | Recipient email addresses and the content of the message | All | United States |
| Functional Software, Inc. (Sentry) | Error monitoring. Session replay is not enabled | Error reports and technical diagnostics, which can include the URL and user identifier involved in a failure | All | United States |
| Stripe, Inc. | Invoicing and payment processing | Billing contact details and invoice amounts. No call data | All | United States |
| Cloudflare, Inc. (Turnstile) | Bot protection on sign-up and sign-in forms | Browser and network signals from the person submitting the form. No call data | All | United States and global edge |
| OpenAI, Google (Gemini) and Perplexity | Answer engines queried by AI Visibility to see how they describe a brand | The brand names and prompts a customer configures. No call data and no personal information about call participants | AI Visibility | United States |
Changes to this list
We give at least thirty days’ notice before adding a provider that will process customer personal information, by updating this page. To be told by email when it changes, write to hello@axoapp.ai. The Data Processing Addendum explains how to object to a new provider.
If you connect your own accounts to AXO.Voice, such as your own telephony provider, that provider acts for you under your agreement with it and is not an AXO sub-processor.