Privacy Policy
This page describes SentientScope's actual data-handling architecture in plain English. It is not yet the final legal policy; the binding version will be published after legal review.
What we store, and what we don't
Your original contracts and SOWs remain in your own document system (Google Drive / Shared Drive, or OneDrive / SharePoint) when linked. SentientScope stores the human-verified commitments extracted from them, the minimum exact excerpts needed to explain why a commitment or alert exists, and source references — not a permanent copy of the full file.
A file uploaded directly is a transient processing input: it is parsed once to build the Living Scope and then deleted, with an automated cleanup enforcing a hard 24-hour maximum for failed or abandoned processing.
Your mailbox is never stored. Only messages forwarded or CC'd to a project address are evaluated, and for project-relevant messages we keep only the identifiers, the minimum relevant excerpt, and the decision history. Approval emails SentientScope itself sent, and the client's responses to them, are retained as workflow records.
AI processing
AI is used to extract commitments, classify candidate events, and draft bounded summaries. Every AI request is built from SentientScope's own records for a single task; no product feature depends on an AI provider retaining prompts, files, or conversation history, and customer content is not used to train models.
Security and access
Workspace data is tenant-isolated and access-controlled server-side. Multi-factor authentication is required for administrative roles. Platform staff cannot browse customer content; any future support access would be time-limited, reason-coded, and audited.
Your controls
You can export or delete your structured workspace data, disconnect any integration at any time, and configure retention. Questions: contact us through your workspace.