About BuildZoom Data

How it works

Here’s what goes into maintaining the construction industry’s largest dataset.

1. Identification
   A research team is responsible for updating, maintaining, and prioritizing an index of public data sources, which includes state and license boards, building and planning departments, and county clerk’s offices.  
2. Extraction
   A variety of methodologies are applied to extract data from sources. 
3. Transformation
   Extracted data is stored in a nested JSON format in Postgres before it is structured and schematized via a combination of human and machine interpretation methods. 
4. Mapping
   Data is mapped into the construction graph, allowing for cross-entity relationship building.  
5. Enrichment
   A bundle of ML methods are applied to develop second-order data points (e.g. entity type classifications). Data is additionally aggregated along several axes to support marketplace applications.