One address in. A fully sourced property file out.
PropertyRiskIQ pulls parcel, structure, valuation, permit, flood, wildfire and fire response data in a single pass, and stamps every figure with its provider, its retrieval time and the raw response behind it. Built for regional carriers and MGAs who need the file to hold up long after the bind.
No policy admin integration. No book upload. No loss history leaves your systems.
Property file
48 Halstead Lane, Petaluma, CA 94952
- Year built
- 1968
- ATTOM Data Solutions, retrieved 2026-07-24 14:08 UTC
- Flood zone
- X
- FEMA National Flood Hazard Layer, panel effective 2019-11-26, retrieved 2026-07-24 14:08 UTC
- Hazard potential class
- 3 of 5
- USDA Forest Service Wildfire Hazard Potential 2020, retrieved 2026-07-24 14:08 UTC
- Nearest station
- Station 3, 1.4 mi by road
- Google Places and Distance Matrix, retrieved 2026-07-24 14:08 UTC
7 sources queried in parallel
Completed in 61s
The problem
The property file is still assembled by hand, one browser tab at a time
To underwrite a single property risk, an underwriter opens the county assessor, the FEMA Map Service Center, a wildfire layer, a satellite view, a permit portal, an AVM and a protection lookup, then retypes the findings into the file.
It takes the wrong amount of time
Twenty to forty minutes of clerical assembly per risk, before any judgment is applied. The scarce resource in a small underwriting team is attention, and this consumes it on retrieval.
It is not reproducible
Two underwriters working the same address reach the file differently, because there is no shared definition of which sources are consulted or in what order.
It cannot be reconstructed later
When a claim, an exam or a treaty review asks what was known at bind, a file of retyped values cannot answer. The sources have moved on and nothing recorded where the numbers came from.
How it works
Three steps, and the third is the one that matters
01
Enter the address
Autocomplete resolves it to a single parcel. Nothing else is required: no submission, no book, no policy administration integration.
02
Every source is queried at once
7 data sources run in parallel rather than in sequence, so the file assembles in about a minute instead of an afternoon of tab switching.
03
The file records its own provenance
Each figure lands with the provider that produced it, the moment it was retrieved, and the raw provider response kept alongside it.
Why evidence first
A score you cannot explain is a liability at exactly the wrong moment
Underwriting files are written once and read under pressure: in a rate filing, a treaty review, a market conduct exam, a disputed claim. In each of those rooms the question is never what the score was. It is what was known, when it was known, and where it came from.
Read the argument- Every fact names its provider
- Not a blended index. The specific dataset, so a disagreement is settled by reading the source.
- Every fact carries a timestamp
- A hazard layer accurate at bind may not be accurate at renewal. The file knows which it was.
- The raw response is retained
- The provider payload stays attached to the analysis, not just the value we parsed out of it.
- Gaps are shown as gaps
- Where no live source is connected, the file says so rather than filling the space with a plausible number.
What is on the file
Seven layers, queried together, reconciled to one parcel
Each layer comes from a named source. The full provenance table, including the layer that is not connected yet, is published.
Structure and parcel
Year built, living area, lot size, construction type, roof cover and the feature flags your eligibility rules key on.
Valuation and permits
Automated valuation, full sale history, assessment and tax, and the permit record that tells you what has actually been done to the building.
Flood
FEMA zone and subtype, Special Flood Hazard Area status, base flood elevation with its datum, and the effective date of the governing panel.
Wildfire
Wildfire hazard potential class and annual burn probability from the published USDA Forest Service raster, with matched mitigations.
Fire response
Responding stations within five miles, road distance and drive time to each. Reported as distance, never dressed up as a protection class.
Aerial imagery
Satellite view at neighbourhood, property and detail zoom, captured against the same coordinates as every other layer on the file.
Adoption
Nothing to integrate, so nothing to wait for
PropertyRiskIQ takes an address and returns a file. It does not connect to your policy administration system, it does not ingest your book, and it does not need your loss history. That is a deliberate design constraint, not a missing feature.
No book upload
The product never holds your policy or claims data, which keeps it outside the review that a system holding your book would trigger.
Data flows one way
Reference data flows from providers to you. Nothing flows the other direction.
Evaluable in an afternoon
An underwriting manager can run real addresses against files they already know the answers to, and judge it on those.
Fits beside what you have
It complements imagery scoring and rating tools rather than replacing them. The record it builds is the layer underneath both.
Questions
Frequently asked
- What is PropertyRiskIQ?
- PropertyRiskIQ is property intelligence software for property and casualty underwriting. You enter a property address and it assembles a single underwriting file covering parcel and structure detail, valuation, permit history, FEMA flood zone, wildfire hazard potential, nearest fire response and aerial imagery. Every figure on the file records which provider it came from, when it was retrieved, and the raw provider response behind it.
- Who is it built for?
- Regional carriers, MGAs, MGUs and program administrators writing homeowners and small commercial property. The typical team has between three and thirty underwriters and assembles property facts by hand today, across the county assessor, the FEMA Map Service Center, a wildfire layer, satellite view and a permit portal.
- How is this different from an imagery-based property score?
- Platforms such as Cape Analytics and ZestyAI infer property condition from aerial and satellite imagery and return a proprietary score. PropertyRiskIQ solves the step before that: assembling the sourced record across public and commercial datasets and showing its provenance. The output is a cited file rather than a score, so an underwriter, a reinsurer or a regulator can see exactly what was known at bind time and where each fact came from. The two approaches are complementary.
- Do we have to integrate it with our policy admin system?
- No. PropertyRiskIQ takes an address and returns a file. There is no policy administration integration, no book upload, and no requirement to send us your submission data or your loss history. That is deliberate: it keeps the product usable without a vendor risk review of a system that holds your book.
- How long does one property take?
- Every data source is queried in parallel, so a lookup completes in roughly a minute rather than the time it takes to open and reconcile a dozen browser tabs. The exact time depends on how quickly the underlying provider services respond.
- Where does the flood determination come from?
- Flood data is queried live against FEMA's National Flood Hazard Layer. The file records the flood zone and subtype, whether the property falls inside a Special Flood Hazard Area, the static base flood elevation and its vertical datum, and the effective date of the map panel. For high-hazard zones it also retrieves the FIRMette panel. This is a data lookup and is not a certified Standard Flood Hazard Determination.
- Is PropertyRiskIQ generally available?
- Not yet. It is running as a working product with live data sources, and we are onboarding a small number of design partners who want the evidence file in their own underwriting workflow. Pricing and terms for that program are on the pricing page.
- Does PropertyRiskIQ make the underwriting decision?
- No. It assembles and cites the evidence. Risk selection, pricing, and any decision to bind, decline or refer remain with the underwriter. The product does not produce a bindable rate and is not a rating engine.
Bring an address you already know the answer to
The fastest way to judge this is to run a property from your own book and see whether the file matches what your team found by hand, and how long each took.