Promptev, Inc. ("Promptev") uses the following Sub-Processors to assist in providing the Services described in our Terms of Service and any signed written agreement between Promptev and the customer. This list is maintained in accordance with our Data Processing Agreement.
We will provide at least thirty (30) days’ prior written notice before engaging any new Sub-Processor. Customers may subscribe to updates by contacting privacy@promptev.ai.
| Sub-Processor | Purpose | Location |
|---|---|---|
| Google Cloud Platform | Cloud infrastructure, compute, storage, database hosting, in-memory cache and task queue | United States (us-east4) |
| Cloudflare | DNS, CDN and TLS termination for the promptev.ai website and api.promptev.ai. The Promptev application is served directly from Google Cloud and does not route through Cloudflare | United States |
Each AI function of the Services can run on the customer’s own provider or on Promptev’s. A function the customer has not configured with its own provider runs on the Promptev Sub-Processor listed below for it. Configuring one function (for example, your own model for answers) does not change the others (document reading, extraction, embeddings).
| Sub-Processor | Purpose | Location |
|---|---|---|
| OpenAI | Embeddings for knowledge search, and extraction from documents, when the customer has not configured its own embedding or extraction model | United States |
| Google Cloud (Vertex AI) | Reading scanned pages and images (OCR), when the customer has not configured its own vision or OCR model | United States |
| PromptevAI model providers: DeepSeek, Anthropic (Claude), OpenAI, NVIDIA and others listed here when added | PromptevAI: the models promptev provides, for answers from agents that use a PromptevAI model, when the customer has not connected its own model. From November 22, 2026 | United States; People’s Republic of China (DeepSeek) |
Where the customer configures its own provider for a function, that provider is the customer’s own processor under the customer’s own agreement with it, and Promptev’s Sub-Processor is not used for that function.
Single sign-on is configured by each customer against their own identity provider (for example Microsoft Entra ID or Google Workspace). That provider is the customer’s, not a Promptev Sub-Processor.
| Sub-Processor | Purpose | Location |
|---|---|---|
| Stripe | Payment processing, subscription billing | United States |
| Sub-Processor | Purpose | Location |
|---|---|---|
| Microsoft (Microsoft 365 / Graph) | Transactional email delivery — sign-in links, invitations, approval and notification mail | United States |
| Slack Technologies | Internal error logging and alerting | United States |
| Sub-Processor | Purpose | Location |
|---|---|---|
| Google Analytics | Website usage analytics (anonymized) | United States |
| Microsoft Clarity | Website session recordings, heatmaps and conversion-funnel analytics on promptev.ai. Form inputs are masked and not recorded | United States |
| Date | Change |
|---|---|
| October 23, 2026 | AI model providers listed. OpenAI (embeddings, extraction) and Google Cloud Vertex AI (OCR) process documents for workspaces that have not configured their own providers for those functions; the September 10 entry wrongly said no model provider was used. Notice given of the PromptevAI model providers (DeepSeek, Anthropic, OpenAI, NVIDIA), from November 22, 2026, for PromptevAI. |
| September 14, 2026 | Microsoft Clarity added for website session recordings, heatmaps and funnel analytics on promptev.ai. |
| September 10, 2026 | Infrastructure moved from Amazon Web Services (us-east-1) to Google Cloud Platform (us-east4); Redis Labs replaced by Google Cloud’s managed cache. Cloudflare added for DNS, CDN and TLS on promptev.ai and api.promptev.ai. Transactional email moved from Postmark to Microsoft 365. Auth0 removed — single sign-on is configured against each customer’s own identity provider. The “platform-managed” AI model provider list removed: model calls are made on the customer’s own provider credentials, so no model provider is a Promptev Sub-Processor. |
| December 22, 2025 | Initial publication |
Questions about our Sub-Processors? Contact us at privacy@promptev.ai.