DataStore

The AI queries your data. It never swallows it.

Upload your sources once. Every canvas queries them where they live — cited back to the exact page — while your corpus stays yours and never becomes anyone's training data.

01 — Ingest

Ingest once. It stays yours.

Drop in PDFs, spreadsheets, transcripts and notes. Each one is parsed and indexed the moment it lands — no manual tagging, no re-uploading into every new chat. The AI reads it where it lives; nothing is copied off into a model.

A private document library being indexed into page-level evidence and cited answers.Ingest once

02 — Auditable

Every answer shows its work.

Ask anything across the whole corpus and the answer comes back with citations to the source — line by line. No black boxes, no guessing where a number came from. You can click straight through to the page it was pulled from.

A verified answer connected to highlighted evidence in a report, spreadsheet and indexed document.Page-level evidence

03 — Reuse

One source of truth. Every workspace.

Connect the same DataStore to any canvas across the org. Update it once and every workspace gets the new answer — no forked copies quietly drifting out of date. Export the whole store and walk away with it whenever you like; it was always yours.

One private indexed corpus supplying relevant evidence to three separate query workspaces.One source · many workspaces

Stay in touch

Not ready to sign up yet?

Leave your email and we’ll send a note when new features ship. No spam.

We care about your privacy. See how we handle your data in our Privacy Policy.

Build on a laptop. Browse, preview, and share from anywhere.