Property data for insurance underwriting and risk modelling
Built for underwriting analysts, catastrophe and exposure modellers, reinsurance treaty teams, climate-risk groups, and pricing analysts at Managing General Agents. Houski ingests hundreds of public, government, and crowdsourced property data sources, then does the hard part: cleaning, conforming, and standardising every field onto one schema across every Canadian province. 19M+ Canadian addresses, one set of column names, one set of units, one set of property type definitions. Useful for portfolio risk analysis, quote pre-fill with verification, and workflows that tolerate or correct individual-property gaps.
A note on accuracy - be honest about what this is
Our data comes from public records, municipal datasets, and community contributions. It is not an appraiser-grade dataset for any single property. Coverage and accuracy vary by jurisdiction and field. For high-stakes single-property decisions like binding an individual policy, use customer-confirmed inputs or a human in the loop.
Houski shines at scale: across thousands of properties, individual errors wash out and signal dominates. See field-level coverage in our fields documentation.
What insurance teams actually do with our data
- Portfolio catastrophe and exposure modelling - Aggregate flood and wildfire exposure, materials, and year-built distributions across a book. Individual gaps wash out, and distribution shape matters.
- Quote form pre-fill (with verification) - Populate a quote from an address, then have the customer or agent confirm. Faster quotes, lower abandonment.
- Market segmentation and prospecting - Find neighbourhoods with older roofs, specific construction types, or no recent permits, then target offerings there.
- Listing-anchored valuations - For properties with recent listing activity, our model is grounded in the most recent list price plus structural details from the listing. The single-property case where Houski is most reliable.
- Change detection via permits - Permits flag properties whose insured value may be stale (renovation, addition, rebuild). Use it to trigger re-underwriting, not to set the value directly.
- Climate risk scoring inputs - We sell the address-level property and geographic attributes your flood, wildfire, or hail models depend on. We don't sell a packaged risk score.
Why insurance teams pick Houski
- Standardised nation wide: Every Canadian province, every municipality, one schema. Field names, units, property type vocabulary, and address formats all normalised to a single definition. No juggling a dozen assessor APIs. Per-field coverage varies by jurisdiction.
- The cleaning work is the deliverable: Hundreds of public, government, and crowdsourced property data sources reconciled into one property entity per address. Years of data engineering so your underwriting team stops hand-stitching exports.
- 200+ fields in the schema: Construction, structure, lot, listing history, plus model-filled estimates on the fields where the training data supports it.
- Independent and unbiased: Not affiliated with the real estate industry or government. We pull public datasets. No licensing restrictions.
- Usage-based pricing: Pay per query, $99 USD monthly minimum. See pricing or talk to us about volume agreements.
Frequently asked questions
• What structural attributes do you have?
Materials, roof type, year built, square footage, stories, garage and basement, lot dimensions, and more. Coverage varies by jurisdiction. Our fields documentation lists every field and its coverage.
• Do you have flood or wildfire risk scores?
No packaged risk score. We sell the underlying property attributes and geographic data your risk models depend on.
• How fresh is the data?
Daily updates. New listings, AVM estimates, permit filings, and assessment changes reflect within 24-48 hours where source data updates frequently.
• Can we use this for individual quote pre-fill?
Yes, with verification: pre-fill from our /properties endpoint, then let the customer or agent confirm or correct. Don't bind on un-verified pre-fill alone.
• Can Houski alone set a premium?
Not for an individual policy. For portfolio pricing, segments, or rate-class assignment at aggregate scale, Houski works well. For binding a specific policy, use customer-confirmed inputs or a human underwriter.
Ready to talk?
We work with insurance carriers, MGAs, brokerages, and insurtech startups across Canada. Reach out and we'll walk through how Houski fits your underwriting or risk pipeline.