Our Data Pipeline

We ingest deed records, property valuation data, owner names, lot size, building square footage, and last-sale dates from county assessor and tax collector offices across all 50 US states. This forms the backbone of our property registry.

We source official parcel boundary polygons from government GIS databases and maintain them in a PostGIS-enabled spatial database. This powers our polygon draw tool, allowing you to select custom territory shapes and retrieve all parcels within.

We integrate high-resolution satellite imagery via Google Maps Static API to pull overhead shots of any property. Our AI vision models then run on these images to detect conditions: roof age, lawn quality, pool presence, impervious surface, and more.

Google Street View imagery provides ground-level context for properties. Our vision pipeline references both aerial and street-level perspectives to make more accurate property condition assessments and to anchor AI render generation.

Anthropic Claude Vision processes property imagery to assign quality scores, detect service opportunities (e.g. "lawn needs work", "roof shows age"), and flag high-intent prospects for your specific vertical. This is proprietary and not sourced from a third party.

Before any postcard is dispatched, each mailing address is verified against the USPS National Change of Address (NCOA) database through our Lob integration. Invalid or moved addresses are flagged to prevent wasted spend on undeliverable mail.