Skip to main content
The Documents page collects everything: files synced from Google Drive, OneDrive and SharePoint, Box, Dropbox, and Procore, attachments from connected email, mail sent to the project inbox, files dropped into a chat, and anything you upload directly.

​
Browsing

Documents live in a two-pane explorer — a folder tree on the left, files on the right — scoped to your active project, with a running document count in the header. The file list shows Name, Type, Category, State, Source, Version, and Last updated. Long lists page in with Load more.
  • Search files matches on name.
  • Filters narrows by All types, All stages, All sources, and an Added after / Added before date range. The button carries a badge with how many filters are active; Reset clears them.
Folders mirror how your files are organised in the tools they came from — there’s nothing to create or rename here.

​
Uploading

Click Upload files, then drag files into the dropzone or click to browse. Before uploading, pick who can see them under Visibility:
  • Everyone in the organization
  • This project
  • Only me
Click Upload.
Large files are handled for you. Big drawing sets and reports upload directly to storage in a way that survives a dropped connection — a transfer that stalls picks up where it left off instead of starting over. There’s nothing to switch on; just be aware that a very large file takes a while, and large uploads are sent one at a time so they don’t saturate your connection.
P6 schedule exports (.xer) uploaded here are routed into the Schedule pipeline instead of the document library. Select a project first.

​
What happens to every document

Each file goes through the same steps, whether uploaded or synced:
  1. Stored and versioned — the same file arriving twice never creates a duplicate, and each change becomes a permanent version you can always go back to.
  2. Parsed — PDFs (with OCR for scans), Word, Excel, images, and email formats, preserving page, line, and position.
  3. Classified — into construction document types: drawings, specifications, submittals, RFIs, contracts, change orders, invoices/pay applications, schedules, meeting minutes, daily logs, safety docs, permits, closeout docs, and more.
  4. Extracted — key fields pulled from contracts, drawings, submittals, invoices, RFIs, and change orders, each anchored to the exact spot in the file it came from.
  5. Indexed — made searchable so the assistant can find and cite it, down to the page and line.
Parsing runs in the background after upload. A document’s State column reads Active or Superseded; each individual version carries its own parse status (Pending, Parsed, Failed, Skipped size, Skipped type) in the Versions list.

​
The document drawer

Click any document to open its details panel. Actions along the top: Open (in the viewer), Download, Parsed text (shows the text Brink extracted — reads No text when there isn’t any, and Loading… while parsing is still running), Supersede…, and Delete. Below that:
  • Details — type, source, visibility (which you can change here), state, and document ID.
  • Classification & organization — what Brink classified it as and how confident it was, plus stage, category, division, and the organized path. Documents that haven’t been through the pipeline yet say so.
  • Extracted fields — each field with its value and a p. / L. jump link back to the exact page and line it was read from. Low-confidence extractions carry a Review chip with a confidence percentage.
  • Versions — the full version history, newest first, each with its origin, timestamp, and parse status; the live one is tagged Current. Drag a new revision onto the Add version dropzone to file it against the same document.
  • Upload history, Supersession (what supersedes this, and what this supersedes), and Changes between versions — a summary of what changed and how significant it was.
  • Provenance — origin, when it was ingested, the file’s checksum and size, and which parser read it.
  • Activity — the audit trail of every action taken on the document.
Deleting a document hides it everywhere but keeps it in the audit trail.

​
Deleting folders

Owners and admins can clear out whole folders: click Select in the folder pane, tick the folders, and click the trash button.
The confirmation dialog shows exactly how many documents will be removed and asks you to type DELETE. Every document inside those folders and their subfolders is removed from search, chat, and the Document Center. The files themselves are kept, so an administrator can reverse it.

​
The viewer

Opening a document — or clicking any citation to it from chat — opens the in-app viewer at the exact cited page, with the referenced region highlighted and briefly flashed so you can see what the answer was reading. Multi-page documents get Prev / Next and a page indicator. Images preview inline. File types Brink can’t render say so — “Preview isn’t available for this file type” — and you can download the original instead.