Platform Guide

Datasets & connectors

Datasets are tenant-scoped containers over uploads and live connectors, with scheduled sync, change reconciliation, and permission snapshots that flow forward to every query.

Supported connectors

SharePoint, Google Drive, OneDrive, Dropbox, Notion, Slack, email, S3, databases, and websites. Each connector reports its own capability manifest, so the platform knows what it can and cannot do — incremental sync, deletes, permission reads — rather than assuming a lowest common denominator.

Permission snapshots

A connector does not only bring content across; it brings the source system's access rules with it. Those snapshots are what let retrieval filter by source ACLs at query time, so a document a person cannot open in SharePoint is not something an answer can quote to them.

Change reconciliation

Scheduled syncs reconcile against what was previously indexed rather than re-ingesting wholesale: edits update in place, deletes propagate, and moved documents keep their identity. That stability is what makes citations durable — an evidence pointer taken last quarter still resolves.