Skip to main content

​
Inspection Reviewer (Report Workspace)

The Inspection Reviewer is the report-finalization workspace on the web admin. It’s where you walk through every finding the inspector captured on mobile, polish narratives, run AI Draft for any field-noted findings, edit photos, organize the summary, generate the PDF, and publish to the client. This article documents the layout, the workflows, and the AI assistance that lives here. For the publish step (what happens after you click Publish), see Reports & Publishing. For the onboarding tutorial, see Generate Your First Report.

​
When you use it

  • As soon as the inspector marks an inspection complete on mobile — the inspection becomes available for review
  • For multi-day inspections — review what’s been captured so far without waiting for the final completion
  • For edits after publish — re-open a published inspection to fix something and re-publish
You reach the Reviewer from Inspection Details → the Open in Inspection Reviewer button on the Inspection Details tab. The URL is route-scoped to the inspection ID.

​
The 4-column layout

[SCREENSHOT: inspection-reviewer-4col.png — desktop, the Inspection Reviewer showing four columns: a scrollable Findings list on the left, a Photo carousel in column 2, a large Narrative Editor in column 3, and AI Suggestions in column 4.] The Reviewer is built as a 4-column grid optimized for one-finding-at-a-time editing: The columns resize responsively. On narrower viewports (laptop screens), columns 3 and 4 collapse into tabs.

​
Column 1 — Findings list

The leftmost column shows every finding for the inspection, grouped by section. Each finding card displays:
  • A small thumbnail of the photo (or video icon)
  • The caption
  • An IN SUMMARY badge if flagged for the executive summary
  • A small status dot:
    • green — narrative complete, ready to publish
    • amber — field note exists but narrative not yet written (AI Draft candidate)
    • red — flagged as needing review
Click a finding card to load it into columns 2–4. The current selection is highlighted with a left border.

​
Filtering the list

A small filter bar at the top of column 1:
  • All / Summary only toggle — same idea as the mobile Review screen’s “Show summary items only”
  • All / Has Field Note filter — surfaces the AI Draft candidates
  • Section dropdown — restrict to one section

​
Column 2 — Photos & metadata

The middle-left column displays the active finding’s media and metadata. If the finding has multiple photos (currently rare — one photo per finding is the norm), arrows scroll through them. Tap any photo to view full-size in a modal with zoom. If the finding is a video, a thumbnail with a Play overlay; click to open the video player at /player.

​
Markup view

If the photo was annotated on mobile with Photo Markup, the annotations appear baked into the photo. The original-unannotated version is not retained — the inspector’s annotated version is what the report renders.

​
Metadata panel

Below the photo: All edits auto-save to Firestore as you type (debounced).

​
Column 3 — Narrative Editor

The rich-text editor for the active finding’s narrative. This is where you spend most of your time. [SCREENSHOT: narrative-editor.png — desktop, the Narrative Editor in column 3 with a Tiptap toolbar (bold, italic, underline, bullet, table) and the finding’s narrative below.]

​
Editor capabilities

Built on Tiptap (same engine as the Communications email templates and contracts):
  • Bold / Italic / Underline
  • Headings (h2, h3 within a finding — rarely needed)
  • Bullet and numbered lists
  • Tables — for itemized recommendations or fee schedules
  • Inline images — embed reference photos within the prose (use sparingly — most photos belong on the finding card, not in the narrative)
  • Hyperlinks — link to external resources (manufacturer pages, code references)

​
Auto-save

The editor uses a 1.5-second debounce — stop typing for 1.5 seconds and the content commits to Firestore. The save status indicator (next to the editor title) shows: You can leave the page mid-edit; auto-save catches the last keystroke before navigation.

​
Save indicator

A lastSavedNarrativeRef tracks the previously-saved value so the auto-save no-ops when nothing has changed. This avoids spamming Firestore with identical writes.

​
Column 4 — AI Suggestions

The right column surfaces AI-generated suggestions for the active finding. These are different from the Mobile AI tools — the Reviewer’s AI assistance is geared toward refinement rather than initial generation.

​
Suggestion types

Click any suggestion to load it into the narrative editor (preview before applying).

​
Cost

  • Alternative phrasings: polish-narrative = 2 IT each
  • Library matches: semantic-search = 1 IT each
  • Field-note draft: generate-narrative = 2 IT
  • Polish suggestions: bundled with the in-editor experience
See AI Features Guide for the full operation reference.

​
AI Draft — the batch workflow

When inspectors capture a finding on-site without writing the full narrative — they jotted just a field note — the AI Draft tool batches the narrative generation for the office. Access AI Draft from the Findings list filter: select Has Field Note. The Reviewer surfaces a Draft All button at the top of the column. Clicking Draft All:
  1. Queries every finding in the inspection that has a field note AND no narrative
  2. Generates a CAR-format narrative for each (generate-narrative = 2 IT per finding)
  3. Populates each finding’s narrative field — you can review and edit individually
This is the operator’s main labor-saver — inspectors capture findings fast on-site, the office polishes back at the desk.

​
System Performance Opinions (SPOs)

The Reviewer also surfaces the SPO editor for each major system. From the column 1 view, scroll to the section header to see the section’s SPO. Click the SPO card to open the editor:
  • Edit manually (free text)
  • Or click Generate to have the AI produce a section-level synthesis (generate-spo = 10 IT)
SPOs appear in the executive summary on the report. See AI Features Guide → SPO for the full mechanics.

​
Executive Summary

A dedicated tab or section (depending on the template) collects the findings flagged Include in summary. Use this to:
  • Review the order — drag-reorder findings in the summary section
  • Edit summary-specific copy (intros, footers, scope notes)
  • Run AI Priorities — same flow as mobile’s Review Screen, ranks summary findings by severity

​
Generate PDF preview

A button (typically top-right) opens the PDF preview in a new tab. The PDF is rendered server-side from inspection data + template + branding. If the preview surfaces issues (page breaks in odd places, photos cropping unexpectedly), use the Templates Overview → Canvas tab to adjust template-level layout, then regenerate. The PDF preview is not stored permanently. Regenerate after every edit.

​
Publish

When the report is ready, click Publish (typically a top-right button). The Publish action:
  1. Generates the final signed PDF and stores it in Google Cloud Storage
  2. Sets reportStatus = 'published' on the inspection
  3. Fires report.published — your communications workflow takes it from there (client email, agent email, etc.)
  4. Records a permanent entry in the History tab on the Inspection Details page
Published reports remain editable. Re-publishing increments the version number; old versions stay accessible in the Reports tab on Inspection Details. See Reports & Publishing for the publish-side workflow details.

​
Keyboard shortcuts

The Reviewer is designed for fast keyboard-driven editing:

​
Common errors

​
What this page does NOT do

  • It does not send the report email. That fires from the workflow engine after Publish.
  • It does not regenerate the report PDF on every edit. Click Generate PDF Preview manually when you want to see the rendering.
  • It does not delete findings. Use the Inspection Details page’s Inspection Details tab or the mobile Review screen.
  • It does not change inspector assignment. That’s Inspection Details.