Sub-Processors

Purpose: Application hosting, serverless functions and scheduled jobs for the operator portal, widgets and webhook endpoints, plus file storage for business logos and for photos and files people send through the chat, the form, text and email.

Data: Operator and End-User data in memory during request handling; stored logos, photos and attachments.

Purpose: Managed Postgres database (hosted on AWS in the U.S. East region) for account data, knowledge-base content, conversations, bookings, CRM records and encrypted credentials.

Data: Persistent storage of Operator and End-User personal data.

Purpose: Managed Redis used for rate limiting on sign-in, public forms, the widgets and incoming webhooks, and for de-duplicating incoming email.

Data: IP addresses, the email addresses used in sign-in attempts, and incoming-email message IDs, held as short-lived keys (minutes to 24 hours).

Purpose: Large language model API for replies to chats, texts and emails, conversation summaries, escalation briefs and suggested replies, the Owner Agent, review reply drafts, campaign email templates, reading photos customers send, sorting new email at a forwarded inbox, checking lead conversations for signs of a scam, and quality checks on call transcripts.

Data: The inputs a feature needs: Operator settings and knowledge-base excerpts, the conversation in progress, relevant contact and booking details, call transcripts, review text, photos customers send, and the sender, subject and text of new email at a forwarded inbox. Not used to train Anthropic models.

Purpose: Voice agent for AI Receptionist: speech recognition, voice generation, and call transcription for inbound calls. ElevenLabs keeps its own recording and transcript of each call.

Data: Call audio, call transcripts, caller phone numbers, and the business context the agent answers from.

Purpose: Business phone numbers, routing inbound calls to the voice agent, missed-call detection, and sending and receiving SMS and MMS.

Data: Phone numbers, SMS content and attached media, call and delivery metadata.

Purpose: Sending and receiving email: notifications and alerts, booking confirmations and reminders, conversations by email, email a business forwards from its own inbox, and AI Loyalty Manager and AI Prospector campaigns.

Data: Recipient and sender email addresses and names, message content and attachments, delivery and engagement metadata (sent, delivered, opened, clicked, bounced, unsubscribed).

Purpose: Checking that AI Prospector prospect email addresses can receive mail before a campaign emails them.

Data: Prospect email addresses.

Purpose: Subscription billing, top-up pack charges, payment-method storage, invoicing, tax calculation, fraud screening.

Data: Operator billing identifiers, payment-method metadata (Stripe holds the card; we hold a token), invoice history. We do not store full card numbers.

Purpose: Syncing reviews and publishing the review replies the Operator approves; reading calendar availability and events (to show them in Velogics) and managing the appointment events Velogics books. See Privacy Policy § 11.

Data: Encrypted OAuth tokens; reviews and replies; availability, calendar events and appointment details.

Purpose: Outlook / Microsoft 365 calendar integration via the Microsoft Graph API, reading availability and calendar events and managing appointment events on behalf of the connecting Operator.

Data: Encrypted OAuth tokens; availability, calendar events and appointment details (times, attendees, booking details).

Purpose: Contact sync with the Operator's HubSpot account, when connected.

Data: Encrypted access token; contact details and lead information.

Purpose: Lead, appointment and note sync with the Operator's Follow Up Boss account, when connected.

Data: Encrypted API credentials; lead contact details, appointments and conversation notes.