MadMantra

Sub-processors

Last updated: May 2026. Third-party services that process customer data on MadMantra's behalf. Required disclosure under GDPR Article 28 and most enterprise Data Processing Agreements.

We use a small set of vetted service providers. Each one is bound by its own data-protection terms, and we sign Data Processing Agreements with every sub-processor that handles personal data. We notify customers in advance of material changes to this list per the terms of our DPA.

ServicePurposeData processedRegion
Anthropic ↗AI model inference for chat, agents, and content generationChat messages, prompts, agent context (company info, tasks, documents)United States
OpenAI ↗AI model inference for selected tasks (audio transcription via Whisper)Voice notes uploaded via WhatsAppUnited States
Google (Gemini / Veo) ↗AI video generation for ad creativesAd-creative prompts and generated outputsUnited States
Stripe ↗Payment processing, subscriptions, and Connect payoutsCustomer billing email, payment method (handled by Stripe — we never see card numbers), subscription stateUnited States
Resend ↗Transactional + commercial email sending and inbound email parsingRecipient email addresses, subject lines, message bodies, delivery eventsUnited States
Neon ↗Managed Postgres for the platform database and per-customer databasesAll persistent customer dataUnited States (us-east) — EU region available on request
Render ↗Application hosting for the platform API, worker, and per-customer deployed servicesApplication code, environment variables, runtime logsUnited States
Cloudflare ↗CDN, edge proxying, DNS, DDoS protectionRequest metadata (IP address, user agent, path) and cached page contentGlobal edge network
Cloudflare R2 ↗Object storage for uploaded media (WhatsApp voice notes, images, attachments)User-uploaded filesUnited States / Global
Meta ↗WhatsApp Cloud API messaging; Meta Ads delivery on the platform-managed ad accountWhatsApp message content (only for users who connect WhatsApp); ad creative assets and budgetsUnited States
GitHub ↗Source control for each customer’s deployed business app (one repo per company)Generated application source code, commit historyUnited States
Inngest ↗Background job orchestration (nightly agent cycles, scheduled tasks)Job payloads (company IDs, task identifiers, execution metadata)United States
Vercel ↗Frontend hosting and edge functions for the dashboardRequest metadata, analytics page viewsGlobal edge network

Customer-deployed apps

Each customer's business app is deployed to its own dedicated Render service with its own dedicated Neon Postgres database. That deployment is the customer's own infrastructure footprint — they own the code, can export it, and can move providers if they choose. The sub-processors above process the customer's data only in the context of running these services.

Questions

For DPA requests, sub-processor inquiries, or compliance questions, email hello@madmantra.com.