Using Intake Damage Scan

What it does

Intake Damage Scan lives at /app/ai/intake-scan and exists to answer one dispute before it happens: "you damaged my car." At check-in, you drop a photo of the customer's vehicle and the tool maps the pre-existing damage that's already visibly there — scratch, dent, chip, crack, scuff, paint, or other — grouped by severity (severe / moderate / minor) with a location and a short note for each item. It's read-only and advisory: nothing from a scan is ever saved automatically, and the uploaded photo itself is never stored. The scan is a record you review and then attach to the customer's real intake / DVI record yourself through your normal flow, before any tint, PPF, or coating work starts.

Where to find it

There's no dedicated "AI → Intake Scan" sidebar path, and no direct sidebar entry for the "AI Tools" hub either — that hub isn't in the sidebar rail. A generic path in is opening Growth & Marketing from the sidebar (under "Run the shop") and clicking its AI Tools tab, which loads the hub inline (that tab also has an "Open standalone" link straight to /app/ai); Intake Damage Scan is the card in the "Inspect & protect" section, badged "New." Plain Cmd+K opens a different feature ("AI Help," a KB-grounded Q&A popup), while the Cmd+Shift+K page-search palette does list it ("AI · Intake damage scan").

The more reliable, purpose-built way in, though, is from an appointment: open the appointment's calendar page, scroll to the Inspections section, and click the "AI intake damage scan" button there — that's also what carries the appointment context described below.

Running a scan

Drop 1-6 clear, well-lit shots of the panels or angles you want on record into the photo dropzone. The optional "Area" field (e.g. "driver door, front bumper," 80-character max) just labels what you're photographing — it isn't required. The "Scan for pre-existing damage" button stays disabled until at least one photo is dropped; once you click it, it shows a spinner and "Scanning…" while the request is in flight. If the photo can't be read, you'll get a plain error asking you to try a clearer, better-lit shot rather than a silent failure.

Reading the results and the Live / Stub badge

Watch the badge in the top-right of the page: "● Live" means it's hitting the real model; "● Stub mode" means live AI isn't available for your shop (Intake Damage Scan is a Pro feature, and on Pro it drops to Stub once your AI allowance for the period is used up). Below Pro, a scan returns "Your plan doesn't include this AI tool. Requires Pro or higher."; on Pro with the allowance used up, everything you see is sample data (the result's summary note labels it as sample output).

In either mode, the summary card shows a confidence percentage, and — when the photo is too unclear to call — the tool degrades honestly into a low-confidence note telling you to re-shoot or note the damage by hand instead of guessing. If the model reads the photo cleanly and finds nothing, you get a green "no damage found" note instead of an empty list. When it does find damage, items are grouped into Severe / Moderate / Minor badges up top, then listed underneath with a kind badge (Scratch, Dent, Chip, Crack, Scuff, Paint, Other), the location, and a note.

Attaching to the appointment record

If — and only if — you opened this tool from an appointment's Inspections section (so the URL has ?appointmentId=), a damage result that finds at least one item shows an extra button: "Attach to this appointment's DVI record," which deep-links you straight back to that appointment's calendar page at the #inspections anchor. If you opened the tool generically from the AI hub with no appointment context, that button won't appear at all — you'll need to copy the findings into the customer's intake/DVI record yourself. Either way, remember the scan itself saves nothing: the photo is read once for the scan and discarded, so if you navigate away before saving the findings into the real record, they're gone.

Frequently asked questions

Q: What does Intake Damage Scan actually do?

A: At check-in, you drop a photo of the customer's vehicle and the tool maps pre-existing damage already visible in it — scratch, dent, chip, crack, scuff, paint, or other — grouped by severity (severe/moderate/minor) with a location and short note for each item, so you have a record before work starts.

Q: Does it save or store the photo I upload?

A: No. It's read-only and advisory — nothing from a scan is ever saved automatically, and the uploaded photo itself is never stored; it's read once for the scan and discarded.

Q: How many photos can I upload, and is the Area field required?

A: You can drop 1-6 clear, well-lit shots. The optional "Area" field (e.g. "driver door, front bumper") is an 80-character max label and isn't required.

Q: What's the best way to open this tool for a specific appointment?

A: Open the appointment's calendar page, scroll to the Inspections section, and click "AI intake damage scan" there — that's the purpose-built path, and it's also what carries the appointment context needed to attach findings afterward.

Q: What happens if the photo is too unclear for the model to call confidently?

A: The tool degrades honestly into a low-confidence note telling you to re-shoot or note the damage by hand instead of guessing; a genuinely unreadable photo returns a plain error asking for a clearer, better-lit shot.

Q: How do I get the scan's findings into the customer's actual intake/DVI record?

A: If you opened the tool from an appointment (URL has ?appointmentId=) and it found damage, an "Attach to this appointment's DVI record" button deep-links you to that appointment's calendar page. Opened generically from the AI hub, that button won't appear, and you'll need to copy the findings in yourself — either way, nothing is saved automatically, so don't navigate away before saving.