Skip to content

Everything on this page describes the product as it ships today. The full security questionnaire is available on request.

Your data trains nothing

Matter documents, questions and answers stay inside your workspace. They are never used to train or evaluate models.

Every AI action is attributable

Each question, plan, answer and acceptance is written to an append only log with the person behind it.

Privilege is a system property

Ethical walls apply everywhere: search, the library, notifications and AI answers.

Storage
Encrypted at rest. Region pinned on Firm and Enterprise plans.
In transit
TLS 1.2 or higher on every connection, including to reporter APIs.
Model inference
Runs in the same region as storage. Nothing is retained by the model provider.
Retention
Set per matter by an admin. Deletion removes the document, its index entries, cached passages and log references.
Reporter access
Through your own subscription credentials. Only the passage cited is stored.
Access
SSO enforced per workspace. Admins see sessions and devices.
Subprocessors
Listed on the subprocessors page with 30 days notice before any change.

Ethical walls and matter screening

Firm and Enterprise

Screened people cannot see the matter, its documents, its findings or any answer that draws on them.

Audit export

Firm and Enterprise

Per matter export of every action with actor, time and object.

Retention controls

Firm and Enterprise

Per matter retention with a deletion log you can hand to a client.

SSO and session control

All plans

SAML and OIDC. Enforced per workspace, with device and session visibility.

Private deployment region

Enterprise

Storage and inference pinned to a region of your choosing.

Custom citator integrations

Enterprise

Connect the citator your firm already licenses.

A 45 minute walkthrough using a closed matter of yours.

Book a call