Data Review & Approval: The Human Checkpoint Before Quoting

AI extraction is fast but not perfect. The Data Review & Approval workflow ensures a human double-checks every critical spec before a quote goes out — and keeps a clear record of what was approved for each bid.

After analysis completes, the data table opens with all extracted specs organized across nine tabs. An Approved column sits alongside the extraction columns. The Approved values are the single source of truth for vendor quote generation — nothing goes out until you sign off on it.

Nine spec categories

The tabbed editor covers:

1. Project (name, client, location)
2. Codes & Loads
3. Dimensions
4. Bays
5. Roof & Walls
6. Openings
7. Frame
8. Accessories
9. Source (full extraction provenance)

At-a-Glance summary cards

Key building stats appear as summary cards at the top of the data table — width, length, eave height, frame type, and similar. Click any card to jump to that tab immediately. Fields that drive the 3D building model are marked with a cyan indicator so you know which edits will update the visualization.

Approve and edit

Cherry-pick any extracted value with a single click to promote it to Approved. Manually edit any Approved field inline for values the AI got wrong. Changes stage in a batch — a sticky bottom bar shows how many fields have pending changes, with Save and Discard buttons.

Clickable source badges

Every field shows a source badge (e.g., "Spec p.3", "Plan p.1"). Click any badge to navigate to that page in the Plans Room and see the source firsthand.