Roof risk signals, before quote. Filter out the bad ones early.
Underwriting a homeowner's policy without roof data is guessing. We give you size, pitch, age estimate, condition signals, and material likelihood at quote time, so you can route, decline, or surcharge with real information instead of waiting weeks for an adjuster visit.
How it fits in
Day-to-day workflows our data slots into.
Pre-quote screening
Run every inbound quote through our API. Flag roof age over 25 years, complex roofs, or imagery quality signals that suggest manual inspection. Quote turnaround stays fast for clean risks.
Replacement-cost modeling
Square footage, complexity tier, and material guess feed your replacement-cost engine. More accurate ITV calculations without on-site adjuster visits.
Claim triage
Post-storm, run impacted policy addresses through bulk enrichment. Compare pre- and post-storm roof states. Prioritize claims that match damage signal.
Renewal decisioning
Mid-policy renewals can be re-screened cheaply. Roof aged past your underwriting threshold? Flag for non-renewal or roof-condition rider.
Loss-control engineering
Identify policies most at risk of catastrophic loss. Steep-pitch + old roof + heavy canopy = inspection priority. Schedule physical visits where they actually matter.
Fields you'll use most
From the full API response, these are the fields this industry leans on.
| Field | What you do with it |
|---|---|
| roof.area_sqft | Replacement cost is roughly area × material × labor. The headline input. |
| roof.complexity | simple / standard / cut_up / complex / very_complex. Drives a labor multiplier on replacement cost. |
| roof.predominant_pitch | Steep pitch (8/12+) requires anchor systems for the contractor, raising labor cost. |
| property.year_built | Anchor for roof-age estimation when explicit re-roof history is missing. |
| property.estimated_roof_age_years | Underwriting threshold signal. Most carriers flag 20+ years. |
| data_quality.imagery_quality | LOW means we can't see the roof clearly. Route to physical inspection rather than relying on our read. |
| data_quality.confidence | Internal QC flag. LOW confidence reads should be treated as 'unknown' not 'risk'. |
From signup to live
Typical timeline from first call to your first enriched record.
Underwriting + actuarial review
Joint review with your underwriting team. We map our fields to your existing rating engine inputs, identify which signals you act on vs which are nice-to-have.
Compliance + legal sign-off
We provide a model accuracy whitepaper + data lineage documentation. Your compliance team reviews against your state filings.
Sandbox + parallel run
API key for sandbox. Run our enrichment in parallel with your current process for 4 weeks, compare decisioning outcomes. No production decisions on our data during the parallel period.
Production cutover
Once validated, we move into the real-time quote path. Latency is sub-3-second so the customer's quote-form experience doesn't change.
Quarterly review
Carriers running our data review accuracy + decisioning outcomes quarterly. We adjust thresholds + signal mappings as your portfolio grows.
Common questions
Are you a rating bureau or licensed data vendor?
No, and we don't claim to be. We sell raw and derived property data, your underwriting team interprets it. We're not making coverage decisions, your engine is. Most carriers route us through their data-vendor procurement workflow.
Do you replace EagleView or HOVER for claims?
No. EagleView and HOVER ship inspection-grade reports with PDF deliverables for adjusters. We're fast pre-screening data, not human-verified claims documentation. Use us for triage, route to them for confirmed claims.
What's the SLA?
Enterprise tier carriers get a 99.9% uptime SLA + dedicated support. Sub-25k tier customers get best-effort on the public API.
How do we handle PII in the property block?
Pass include=roof,imagery to skip property/owner data entirely. Most insurers can underwrite from roof signals alone without needing the deed-record block. Owner data is opt-in per request.
Quiet, structured, repeatable
We don't sell scoring or decisioning, only the data. Your underwriting models stay yours. We just plug in the missing roof inputs.
Tell us what you're building
We'll review your use case, answer questions, and get you set up if it's a fit.