Find every hail hit.Prove it in seconds.
Upload roof photos. Our computer-vision engine marks every impact, scores severity, and outputs a report adjusters accept on the first review — so you stop losing claims to "no visible damage."
The damage you can't prove is recovery you don't get.
Most claims aren't lost on price — they're lost on evidence. AI Damage Detection puts pixel-precise proof on every line item before the supplement is even written.
"No visible damage" denials kill claims
Adjusters miss 30–40% of impacts on first inspection. Pixel-precise AI markers and severity scores make the damage impossible to wave off.
Manual photo review burns hours per job
Your estimator shouldn't be squinting at 80 phone shots. AI scans every photo in under a minute and ships a marked-up PDF ready to send.
Missed hits cost real recovery
Every undocumented impact is line items your supplement can't justify. Catch them all and the supplement writes itself.
From photo to proof in 3 steps.
Capture
Snap photos from your phone, upload drone imagery, or pull in a folder of inspection shots. Any angle, any resolution.
Analyze
Computer vision scans every shingle for hail bruising, wind creasing, granule loss, and impact patterns — in seconds.
Deliver
Get a marked-up, adjuster-grade PDF with damage IDs, confidence scores, and code citations attached.
Engineered to win re-inspections.
Pixel-precise damage IDs
Every hit gets a coordinate, severity score, and a cropped close-up — not vague "see photo" callouts.
Trained on storm claims
Models trained on millions of inspection photos across hail, wind, hurricane, and tornado claims.
Evidence adjusters accept
Reports include confidence intervals and reference imagery — built to hold up in re-inspection.
60-second turnaround
From upload to finished PDF. No queues, no offshore review team, no back-and-forth.
Works with what you have
Phone, DJI drone, EagleView imagery, CompanyCam exports — point it at any photo source.
One file. Everything linked.
Damage findings auto-attach to your measurement report, estimate, and supplement — one connected job file.
Questions, answered.
How accurate is AI damage detection on roof photos?+
Our computer-vision models are trained on millions of inspection photos across hail, wind, hurricane, and tornado claims. Average per-impact confidence is 94% on standard phone or drone imagery, and every detection ships with a visible confidence score so your team can review borderline calls quickly.
What kinds of photos work — phone, drone, EagleView?+
All of them. Snap with your iPhone or Android on the roof, fly a DJI drone, pull EagleView orthos, or import a CompanyCam album. The model handles low light, oblique angles, and mixed resolutions. No special capture protocol required.
What damage types does the AI identify?+
Hail bruising and impacts, granule loss, wind creasing, lifted shingles, exposed nail heads, mat fractures, flashing damage, soft spots, and storm debris impacts. Every hit is tagged with damage type, severity, and a cropped close-up.
Do adjusters actually accept these reports?+
Yes. Reports include pixel coordinates, severity scores, confidence intervals, and reference imagery — the same evidence format adjusters expect from a field inspector. Contractors using RoofGenius see higher first-review approval rates and fewer re-inspections.
How fast is the turnaround?+
Roughly 60 seconds from upload to finished PDF on a typical 40–80 photo job. No queues, no offshore review team, no waiting on a third-party human reviewer.
Does it tie into supplements and measurements?+
Yes — damage findings auto-attach to your measurement report and AI supplement on the same job file, so you ship one connected document to the adjuster instead of three separate PDFs.
How do I get started?+
Run a free claim first, then sign up for the full platform when you're ready to scale detection across every job.
Catch the damage adjusters miss.
Detect, document, and deliver evidence-grade reports on every storm job.
Get started