The service providers CustomerFlows uses to process customer personal data, what each one does, and where the data is processed.
Last Updated: October 29, 2026
This page lists the sub-processors Keilani Media Group LLC, doing business as CustomerFlows, uses to process personal data on behalf of our customers, as described in our Data Processing Agreement and Privacy Policy.
We give customers at least 30 days' notice by email before a new sub-processor starts processing their data, and update this page at the same time.
| Sub-processor | What it does for CustomerFlows | Data it processes | Where |
|---|---|---|---|
| Cloudflare, Inc. | Hosting, storage, content delivery, security, and our AI features (Workers AI, AI Search, AI Gateway) | All customer data processed by the Service, including messages and AI conversations | Global network; stored data mainly in the United States |
| Supabase, Inc. | Database and sign-in | All customer data stored in the platform | United States |
| Railway Corp. | Hosting for some of our application services | Customer data processed by those services, and their logs | United States |
| Meta Platforms, Inc. | WhatsApp messaging, and the ad conversion and lead features a customer turns on | WhatsApp phone numbers, messages and media; hashed email and phone, click identifiers and conversion values | Global |
| Google LLC | The Google Ads conversion features a customer turns on | Hashed email and phone, click identifiers and conversion values | United States / Global |
| Resend, Inc. | Sending and receiving email | Email addresses and email content | United States |
| Stripe, Inc. | Billing for CustomerFlows subscriptions | The customer's own billing details (no end-customer data) | United States |
| GitHub, Inc. | Runs our encrypted nightly database backup | An encrypted copy of the database while the backup runs | United States |
Questions about our sub-processors: privacy@customerflows.com.
Ready to put this into practice?
CustomerFlows gives you the AI chatbot, trade-ready pipelines, and attribution layer to run the workflow you just read about.