CloudFran Data Processing Addendum (DPA)
Platform Sub-Processors
CloudFran uses the following sub-processors across its platform, including but not limited to the AI Agents, Presentations AI, CRM, CloudFlow workflow engine, and ARC orchestration services:
| Sub-Processor | Purpose | Data Categories | Location |
|---|---|---|---|
| Anthropic, PBC (Claude API) | AI agent reasoning, complex content generation, code generation, RL planning, multi-step orchestration | Prompts, structured context, generated text. No training on customer data per Anthropic's commercial terms. | USA |
| Google LLC (Gemini API) | AI content generation, slide text creation, embedding generation | Prompts, industry parameters, generated text | USA |
| Microsoft Azure | Application hosting, database storage, file storage, compute | All customer data | Central US |
| SendGrid (Twilio Inc.) | Transactional email delivery (share links, verification, notifications) | Email addresses, message content | USA |
| Twilio Inc. | SMS notifications and voice (optional) | Phone numbers, message content | USA |
| Stripe Inc. (incl. Stripe Connect) | Payment processing, marketplace payouts, billing, KYC for connected accounts | Payment data, payout details, KYC information | USA |
Data Minimization Commitment: CloudFran sends only the minimum data necessary to each sub-processor. Anthropic Claude and Google Gemini receive scoped prompts and structured context returning generated text — they do not receive raw customer CRM databases, financial records, or customer lists unless the customer explicitly includes such data in a prompt.
No Model Training on Customer Data: CloudFran has confirmed with both Anthropic and Google that data submitted via their commercial APIs is not used to train their underlying models. CloudFran does not authorize any sub-processor to use Customer Data for training.
CloudFran will provide at least 30 days' notice before adding new sub-processors to this list. The current list is maintained at cloudfran.com/Legal/DPA.