ZONING KNOWLEDGE GRAPH
Traditional zoning databases store entities in isolated tables — parcels in one, codes in another, overlays in a third. ZoningOps stores everything as a property graph. Every parcel is a node. Every zoning code is a node. Every overlay, every jurisdiction, every variance decision — a node. And every relationship between them is a typed, traversable edge.
This means a single API call can answer: “What are all the parcels within 500 meters of a proposed upzone corridor, in jurisdictions that have adopted ADU ordinances in the last 2 years, with an FAR below 1.5?” — traversing relationships in milliseconds, not minutes.
Parcel 04-721-088
MU-4 Mixed Use
Upzone probability: 78%
DATA SOURCES
Full zoning code text and use tables
Parcel geometry and ownership records
State-level overlay and mandate data
Commission calendars and minutes
Authoritative boundary shapefiles
20+ years of code change history
AI LAYER
The graph is the foundation. The AI layer runs on top — trained on decades of regulatory outcomes to deliver predictions, classifications, and anomaly signals that no static dataset can provide.
ML model trained on 12M+ historical rezoning decisions. Scores each parcel 0–100 on likelihood of upzoning within 24 months.
NLP classification of planning commission documents, meeting minutes, and petition filings into structured event types.
Compute maximum by-right development yield for any parcel given its current zoning code, overlays, and height restrictions.
Flags unusual patterns: variance requests inconsistent with neighborhood history, rapid succession of code amendments, outlier setback approvals.
Access the full ZoningOps API. 47M+ parcels. 36,000 jurisdictions. Sub-50ms queries. All via REST and GraphQL.