BlazeDocs was built for AI workflows that handle sensitive documents — legal, medical, financial. The simplest way to protect a document is to never keep it. So we don't.
Your PDF is converted in memory and discarded as soon as processing finishes. We never write your original document to long-term storage — there is no stored copy to leak, breach, or subpoena.
Dashboard conversions keep the Markdown output so you can revisit, search, and chat with your documents. Delete a conversion and the content is removed, including AI analysis and search embeddings.
The /api/v1/convert endpoint returns your Markdown in the response and stores no document content. It is designed for pipelines where retention is unacceptable.
Private Mode conversions skip AI analysis entirely and are automatically wiped shortly after you download the result — nothing lingers in chat history, summaries, or embeddings.
Your documents and their converted output are never used to train models — ours or anyone else’s.
BlazeDocs runs on Vercel and Convex with authentication by Clerk. All traffic is encrypted in transit with TLS, and access to production systems is restricted.
Reviewing BlazeDocs for your team? Email support@blazedocs.io and we'll answer security questionnaires and data-handling questions directly. For the full details on what is stored and when, see how we handle your documents and the privacy policy.