Subprocessors
Twinbook uses a small set of third-party providers ("subprocessors") to operate the service. We keep this list current so partners and privacy reviewers — universities, creators, authors, and institutions — can evaluate it. Customers under a signed agreement receive advance notice of material changes.
Last updated: 2026-06.
| Provider | Purpose | Region |
|---|---|---|
| Google Cloud Platform | Cloud infrastructure — hosting, database, storage, and compute for self-hosted models | United States |
| Cloudflare | API gateway | Global |
| Portkey | API gateway | United States |
| Google Cloud Vertex AI | AI — document understanding, generation, embeddings | United States |
| RunPod | GPU compute — hosting Twinbook's own models | United States |
| Groq | Speech-to-text | United States |
| Google Firebase | Authentication & sign-in | United States |
| Stripe | Payments & billing | United States |
| Brevo | Transactional email | European Union |
| Supadata | Video transcript fetching | Global |
Notes
- Payments. Card data is handled by Stripe (PCI-DSS Level 1); Twinbook does not store raw card numbers.
- Transcripts. For video content, only the public video identifier is sent — no account data, uploaded files, or learning memory.
- No advertising or data brokers. None of our subprocessors are advertising networks or data brokers, and we do not sell or share user data for marketing.
- Training. We do not train or fine-tune models on user content. Twinbook may fine-tune using its own, licensed, public, or synthetic data — never user content — and we use AI providers under terms where submitted content is not used to train their foundation models.
For a detailed subprocessor disclosure (sub-services, data categories, and regions) as part of a DPA or partner evaluation, contact us.