Last updated: June 6, 2026
This document is effective as of the date above and replaces any prior versions. See the "Changes" section below for how notice of material changes is delivered.
This page lists the third-party sub-processors that SaleSage engages to provide the Service. Each entry shows what the provider does, the categories of data it receives, and where the processing happens. We update this page whenever we add, replace, or remove a sub-processor.
1. How we use sub-processors
SaleSage is the data controller for the account, billing, and product-usage data we collect, and the data processor for the meeting content, prompts, and AI outputs that your workspace generates. To deliver the Service we engage the sub-processors listed below. Each sub-processor is bound by a written data-processing agreement, is selected for security and privacy posture, and is reviewed when material changes occur (for example, change in ownership, region, or training posture).
2. Current sub-processors
The table below is the authoritative list as of the date at the top of this page. "Categories of data" describes the broad types of information the provider receives — not every customer or workspace triggers every category.
| Provider | Purpose | Categories of data | Processing region |
|---|---|---|---|
| Supabase | Primary database, authentication, and file storage | Account identifiers, profile data, workspace data, meeting metadata, uploaded files | United States / EU |
| Vercel (incl. AI Gateway) | Application hosting, edge runtime, and AI Gateway routing to model providers | All Service traffic, including transcripts and prompts routed to LLM providers | Global edge |
| OpenAI | Large language model inference for real-time coaching (gpt-realtime, gpt-4o) | Transcript snippets, prompts, contextual metadata, model responses | United States |
| Anthropic | Large language model inference for meeting prep, synthesis, and post-meeting extraction (Claude Sonnet) | Transcript excerpts, meeting metadata, prompts, model responses | United States |
| AssemblyAI | Streaming speech-to-text and speaker diarization | Meeting audio streams and resulting transcripts | United States |
| Recall.ai | Meeting bot orchestration and raw recording capture | Meeting join metadata, raw audio/video recordings | United States |
| SurrealDB | Operational data store for real-time coaching state | Ephemeral coaching state, prompt context, queue state | United States / EU |
| Stripe | Subscription billing and payment processing | Billing contact, plan, invoices, payment status (card data is processed directly by Stripe and never stored on our servers) | United States / EU |
| SendGrid (Twilio) | Transactional email delivery (account, billing, security, product notices) | Recipient email address, message content for transactional emails | United States |
| Upstash (QStash + Redis) | Background job queue and Redis cache for coaching pipelines | Job payloads referencing workspace and meeting IDs, cached lookups | United States / EU |
| Cloudflare Turnstile | Bot and abuse mitigation (CAPTCHA) | Visitor IP address, browser signals, challenge tokens | Global edge |
| Sentry | Application error monitoring and performance traces | Stack traces, request context, scrubbed user/workspace identifiers | United States / EU |
| PostHog | Product analytics, session telemetry, and AI evaluation logging | Pseudonymous user/workspace identifiers, page and event metadata, AI prompt/response telemetry for evaluation | United States |
| Google (Google LLC) | Web and product analytics via Google Analytics, when enabled by configuration | Pseudonymous visitor identifiers, page and event metadata, device and approximate location signals | United States / EU |
| Amazon Web Services (AWS) | Underlying cloud infrastructure (compute, storage, networking) for several sub-processors listed above, including the regions Supabase and Upstash operate in | All Customer Content and account data that traverses or rests on those sub-processors necessarily traverses or rests on AWS infrastructure | United States / EU |
3. Changes and notice
We will post any addition or replacement of a sub-processor to this page at least fourteen (14) days before the new sub-processor begins processing Customer Content. Workspace owners on plans whose order form or data-processing agreement provides for direct notice will additionally receive an email before the change takes effect, giving the workspace owner time to object on reasonable grounds.
Anyone — including customers on self-serve plans, prospective buyers, and security teams — may subscribe to receive an email whenever this list changes. To subscribe, email legal@trysalesage.com with the subject line "Sub-processor notifications" and the address you would like notices sent to. To unsubscribe, reply to any notice or email us with "Unsubscribe" in the subject line.
4. Objecting to a sub-processor
If you reasonably object to a new sub-processor on data-protection grounds, contact us within the notice window described above. We will work with you in good faith to provide an alternative (where one exists) or to scope the processing so the objection is addressed. If we cannot reach a workable resolution and the new sub-processor is essential to deliver the Service, your sole remedy is to terminate the affected portion of the Service in accordance with your order form.
5. Historical changes
Material changes to this list are tracked in the SaleSage release notes and reflected in the "Last updated" date at the top of this page. Prior versions are available on request via the contact below.
6. Contact
Questions about a sub-processor, a regional posture, or a specific data-processing agreement should be directed to legal@trysalesage.com.