For insurance and risk teams

Price the property on facts, not on what the customer remembers.

Pull wall type, roof, age, floor area, flood and environmental context, rebuild cost and occupancy for any UK address before the customer reaches question three.

See what you can do
No credit card · 500 free credits / month
AI assistantPropalt connected
Pre-fill a home quote for 14 Beaumont Gardens: walls, roof, age, flood and rebuild cost.
30+ risk attributes returned in 104ms, before question three. Quote pre-filled:
QUOTE PREFILL● LIVE
14 Beaumont Gardens
London · SW19 4AT
Walls
Cavity · 1930s
Flood
Very low
Rebuild est.
£286k
Occupancy
Owner-occupied
30+ attributes · ~104ms
29M
properties you can score
30+
risk attributes per address
104ms
per call at quote time
24h
environmental data refresh
See it in action
Pre-fill a quote on real property facts, not customer memory.
Pre-fill a quote on real property facts, not customer memory.
The connector in action

Ask it the way you actually work

Real questions from your desk, answered inside Claude or ChatGPT with the Propalt AI switched on. No dashboards to learn.

ClaudePropalt connected
Is this owner-occupied application actually a let?
Signals suggest a let: company ownership and a landlord holding across three postcodes.
Likely a letCompany owned3-postcode portfolio
ChatGPTPropalt connected
Re-score my book for properties where flood risk changed since inception.
216 policies show a change in environmental risk since inception. Flagged for repricing.
216 flaggedRisk changedReprice at renewal
Context

The Problem We Solve

Home insurance is still priced on what the customer can remember: the year it was built, the wall type, whether the cellar has ever flooded. Most of that is guessed or left out.

So the form gets longer, and more people drop out. The ones who finish are often the least accurate. You end up pricing risk you cannot see, and finding out at claim time.

What changes with Propalt

Outcomes you can rely on

01 · Quote prefill

Replace the questions with the register

Resolve an address and return construction, build-age band, floor area, bedrooms and heating system. Fourteen questions become four, and the four remaining are the ones only the customer can answer.

02 · Rating factors

Verified attributes as model inputs

Bulk retrieval by UPRN for model training and back-testing. Test predictive lift against your existing self-reported factors before committing to a pricing change.

03 · Ground risk

Coal and mining exposure by coordinate

Coal Authority area and non-coal mining hazard checks per location. Screening at quote rather than a manual referral.

04 · Fabric

Fabric, heating age and condition indicators

Wall and roof construction, insulation present, glazing, main heating system and age band. The fabric detail behind thermal-efficiency and water-escape exposure.

05 · Portfolio exposure

Re-score the book against verified attributes

Run the in-force book by UPRN and quantify how far the self-reported attributes have drifted from the register position. Usually the most persuasive first exercise a pricing team runs.

06 · Claims validation

Check the property position at claim

Confirm construction, tenure, EPC and transaction history at the point of claim, with a record date attached, rather than reconstructing it from correspondence.

29M UK properties can be scored on 30+ risk attributes each, with environmental data refreshed every 24 hours and quote-time calls in about 104ms.
For developers
Prefer the technical detail? Endpoints cover 90%+ of insurance workloads; full reference at propalt.ai/product/forDeveloper.
get-property-detail, check-coal-area, check-mining-area, get-crime, get-epc-fabric-by-postcode, get-demographics

Quote on what the property actually is.

Free tier · 500 credits / month · No credit card.

View full API reference
No credit card required · Free 500 credits/month