Is my data used to train AI?

For the authoritative answer on how Vulgate handles your data, please read our Privacy Policy and Terms of Service.

What we do with your content

Your documents and chats power your experience in Vulgate:

Document embeddings — When you upload a file, we generate vector embeddings so neural search can find passages by meaning.
Retrieval-augmented chat — When you ask the AI a question, we fetch relevant passages from your documents and pass them to the language model as context for that single request. The model uses them to write an answer with citations.
Internal usage metrics — We track how the product is used (which features, how often) in aggregate, anonymized form to make Vulgate better.

Chat answers and translation — frontier models from Anthropic, OpenAI, and Google, called through each vendor's API.
Speech-to-text — AssemblyAI.

Each provider is bound by their commercial terms governing how API inputs are processed; see Data privacy and GDPR for the full sub-processor list.

Documents and metadata — Supabase (PostgreSQL) with row-level security keyed to your account / Organization.
Original files — Cloudflare R2 (S3-compatible storage), encrypted at rest.

Read our Privacy Policy — the authoritative description of what we collect, why, and how it's used.
Read our Terms of Service — covers your ownership of uploaded content.
Ask us anything — email info@vulgate.ai.