Privacy

What we do with your library

Last updated 12 August 2026 · plain-language summary first, full terms below

Your documents are yours. We process them to answer your questions and run your soundings. We do not train models on them, we do not sell them, and we do not let another account see them — cross-space soundings report only that a bridge exists.

1 · What we collect

Account details, the documents and annotations you upload or import, the questions you ask, and ordinary service telemetry. We do not buy data about you from third parties.

2 · How your documents are used

Solely to serve your account: retrieval, extraction, soundings and the answers you request. Processing is scoped to your spaces, and cross-space soundings are computed without disclosing text across account boundaries.

3 · Model training

We do not train, fine-tune or evaluate models on your documents, questions or annotations. Where a third-party model provider is used, it is under a zero-retention agreement.

4 · Sharing and sub-processors

A short, named list of infrastructure and model sub-processors, each under contract, each listed with its purpose and region. Changes are announced before they take effect.

5 · Retention, export and deletion

You can export your library, annotations and matrices at any time, and deletion removes documents and derived indexes within a stated window.

6 · Your rights and contact

Access, correction, portability and erasure requests, the responsible entity, and a working address for privacy correspondence.

This page is a designed shell. Have counsel supply the operative text before publishing — the structure, tone and summary-first pattern are the deliverable here.