Vertex PCA FAQ.

You can use ARC whenever you are writing a report that requires the Mid-Level Template. Multi-family and client specific templates are coming soon!

1. Sign in to the web app and click Create report.
2. Choose the report type: Property Condition Assessment.
3. Start typing the site address and click it when it appears. The rest autofills.
4. Assign a project number. This becomes the project title, which is also the Word document's filename. You can delete it and write a custom one.
5. Optionally add a building or property name. Both populate in the report.
6. Upload your files, then click Upload.

Add the client in the report information step. Their details populate in the report automatically.

If the report covers more than one building, add each address in the report information step. There's no two-address limit, and you can remove any address you no longer need. You can also come back and update addresses or site information at any point after the report is created.

Open the report in the web app and drag and drop the new files in, exactly as you did the first time. The assistant picks up the new inputs automatically, so you don't need to download the report again.

• Drag and drop, click to select files, or add whole folders.
• Max file size is 150 MB. Anything larger is flagged and skipped.
• Duplicates are flagged, and the same file is never uploaded twice.
• There's no practical file limit. Some PCAs run 1,000+ files.
• The count at the bottom of the screen lets you verify everything landed.
• Processing takes 1 to 5 minutes, depending on file count and your wifi. The top-right counter shows the percentage.
• At 100%, click Export to download the Word document.

Open the Word document you downloaded from the web app, go to Add-ins, and open the assistant. Sign in if you aren't already. The assistant starts working immediately. For a PCA you land on the cover page with a property summary and its evidence. Two ways to move around:

• Follow cursor (default). The assistant tracks whatever section you're typing in. Use this for most of the report.
• The dropdown at the top. Jump to any section. Best used in the photolog. The assistant jumps to the section you pick, but you still have to move your cursor there yourself.

At the bottom, Writing gives you suggestions and Inputs gives you the photos and documents for the section you're in.

Click Follow cursor. When it lights up blue, it's tracking you. If you're in the photolog past the 140 built-in photo placeholders, cursor following stops working on placeholders you added yourself. The assistant still works. Use the dropdown to change sections.

It's how confident the AI is in what it wrote, given the evidence it found. There are three levels:

• High. Confident in the suggestion. The claims should match the evidence provided.
• Medium. The claims and the evidence should match, but you'll likely have to fill in some information from the evidence provided.
• Low. The system couldn't validate that it found the right evidence, so there isn't enough to be confident in the suggestion.

Inside the suggestion itself:

• Bold text. Confident, backed by evidence.
• [FILL]. The assistant isn't confident enough to write this, so it's asking you to. This is usually where evidence was thin.

A low score isn't a reason to throw the suggestion out. It's a signal to open the evidence and check it before inserting, or to refine the suggestion.

• Refine rewrites it, for example shorter when the evidence is thin.
• Describe a change and submit it: "remove number of parking spaces," or "I know the PSI is 95, find the evidence to support it."
• Check the box when you only want part of a suggestion.
• The counter shows how many times a suggestion has been updated.
• Insert text drops it into the document, or copy it and paste as normal.

Then edit in Word like any other text and fill in the [FILL]s from what you saw on site.

Start with Refine, and prompt the system to find the evidence that backs the claims in the suggestion. If that isn't enough, go to Inputs and browse directly:

• Show photos. Scroll the section's photos and open any for a closer look.
• Show documents. Scroll the section's documents the same way.
• All sections. Open the dropdown in the assistant and choose All sections. Scroll through every input, or search a keyword.

Go to the Inputs tab at the bottom. It shows the inputs for the section you're currently in, ranked by relevance. Switch between Show photos and Show documents, and view documents full-size or as thumbnails.

By default, inputs are ranked by relevance to your current section, which the system sorts during upload. What you see first is what we've determined matters most for what you're writing. You can also:

• Sort newest to oldest, or alphabetically
• Filter by document type
• Search by keyword

Under All sections, relevance ranking doesn't apply, so only newest-to-oldest and alphabetical sorting are available.

Input selection isn't perfect. Two ways to find something:

1. Search it. In the concrete pavement section and the right photos aren't surfacing? Search "truck courts" and everything matching comes up.
2. Browse everything. Switch to All sections to scroll through every input you uploaded, photos or documents.

In the Inputs tab, open the section dropdown at the top and choose All sections. Clear the search bar if you have a term in it, or you'll still be filtered by it. From there, choose Show photos or Show documents and scroll through everything on the report.

Your photolog follows the same order as Section 5, starting with site overview and site improvements. Each section, for example Parking, gets two photo placeholders, and each page holds six photos. If a section doesn't need both placeholders, use the dropdown to move to the next one. The assistant shows the top 10 most relevant photos for whatever section you're in.

Click into a photo box, then:

• Insert photo and caption. Adds the photo and a pre-written caption.
• Insert photo. The photo only.

If a caption isn't what you'd say, edit it like you normally would in Word.

Open any photo in the Inputs tab to read its caption, copy it, zoom in, or read the photo summary.

There are 140 placeholders in the photolog. If you need more, you can add them yourself. The placeholders are a table:

1. Insert two new rows, one for the photo and one for the caption.
2. Copy an existing photo placeholder and caption spot, and paste them into the new rows.

Follow cursor won't track placeholders you added, so use the dropdown to change sections.

• Close the assistant and reopen it. Click the x in the top right corner.
• Reload it. Click the i icon in the top right corner and hit Reload.

• Check the bottom of the Add field data page for a file-too-big flag.
• We don't accept files above 150 MB. You'll be notified when a file isn't uploaded.

• If it hasn't been five minutes yet, wait.
• Refresh the page.
• Force finalize the report. Click the three dots in the top right corner of the web app and select Force finalize.
  • Note: any inputs that weren't processed yet will not be included.