The Problem With Fragmented Data
Four Vendors. Four Schemas.
Four Normalization Pipelines.
The data your models need to understand a property exists — it's just scattered across different vendors with different schemas, different update cadences, and no common property identifier. Joining it is a data engineering problem, not a business problem.
Until now.Without the join
Every Question Costs Weeks
One vendor for permits. Another for valuation. A third for renovation cost. A fourth for condition. Each has a different schema. None share a common property ID. Your data team builds the join. Every time. For every property. At scale this breaks.
With Kukun
One Call. Everything Joined.
Permits, Condition Score, AVM, Renovation Cost & After-Renovation Value, Contractor History, System Lifecycle Forecast, Investment Outlook, Market Intelligence, and the Property Record, joined on any U.S. address in a single API call. Just one schema. One authentication. One bill.
Live Example — Real Data · All Nine Primitives
One Address. One Call.
All Nine Primitives Returned.
This is real Kukun data for a single residential property. Every field below came from a single API call.
10109 Lindelaan Dr, Tampa, FL 33618
Single Family Residence
Built 1972
4,046 sq ft
Hillsborough County
Property Record
3 bd / 3 ba 4,046 sq ft
Owner-occupied · Pool · Brick exterior · Carport (3 spaces)
PICO™ Condition Score
611/850
Property condition scalar blending recently replaced systems (roof, HVAC, windows) against aging interior finishes (kitchen, flooring, paint) — near the neighborhood benchmark of 649.
AVM — Permit-Enriched
$1,170,661
Range: $942K–$1.66M. Assessed value: $574K. Florida assessment caps hold the assessed figure well below market; the AVM reprices to current market plus the permitted improvements.
Investment Outlook (KIO)
+2.55% 5-yr growth
Forecast value in 5 years: $1,200,461. Tampa-St. Pete MSA average: 2.69%. National average: 2.08%.
Renovation Cost & After-Renovation Value
$46,200 roof (2024)
Actual permit-declared cost, plus the after-renovation value it adds — calibrated across 146 categories at ZIP-code resolution (Tampa prices, not national averages).
Market Intelligence
Tampa-St. Pete CBSA 45300
MSA-level demand and risk signals. Benchmark any property against its metro, county, and state. ZIP 33618 growth: 2.55% vs FL average 1.0%.
Contractor History
3 licensed contractors on record
Quality Roofing Inc · JD Air Solutions LLC · Carrollwood Window & Door — each linked to this property by permit, with license number, trade, and status.
Permits — Full History (4 on record)
Sub-permit enrichment surfaces ~40% moremajor residential activity than raw municipal feeds — work recorded at the contractor level that never appears in the public permit roll.
Roof Replacement
Remove existing roofing. Repair and re-nail decking. Install new underlayment and asphalt shingles. Modified bitumen system over flat roof areas.
HVAC Changeout
AC system replacement.
Window Replacement (8 windows)
Replace 8 windows size for size.
Front Entry Addition
New construction — front entry addition.
System Lifecycle Forecast — Remaining Useful Life
Foundation
46 yrs
Roof
24 yrs
Kitchen
6 yrs
Windows
22 yrs
HVAC
16 yrs
Flooring
4 yrs
Interior paint
4 yrs
Exterior paint
3 yrs
What This Means For Your Models
The Join is What Makes The Data Modelable
Individual datasets answer individual questions. Joined data — on a common property identifier, with a common update cadence and a common schema — is what your models can actually ingest.
One schema. No reconciliation
Every primitive uses the same property identifier. Permits join to condition scores join to AVM join to lifecycle. No cross-vendor ID mapping, no date-alignment gymnastics.
Algorithm-ready taxonomy
The same 27 permit labels apply across all 5,600+ jurisdictions. "Roof replacement" in Hillsborough County is the same label as in Los Angeles County. No custom parsing before model ingestion.
Absence is a signal, not a null
A property with no HVAC permit in 20 years in a high-coverage jurisdiction is a risk signal. Every absence answer includes the permit label checked, the date window, and the jurisdiction source.
Algorithm-Ready vs. Human-Readable
Structured For Machines. Not For Reports.
There's a meaningful difference between data a person can read and data a model can ingest. Kukun is built for the latter.
What Most Property Data Looks Like
Permit descriptions in free text: “re-roof w/ 30yr arch shingles” — must be parsed before use.
Different field names and structures per jurisdiction — normalization required per market.
No metadata on absence — a null could mean no activity or no coverage. Impossible to distinguish.
Multiple vendor schemas to join — your data team writes the glue code.
Schema changes break downstream models silently.
What Kukun Returns
27 standardized permit labels across all 5,600+ jurisdictions — ingest directly into feature pipelines.
Single schema for all nine primitives, all markets, all property types.
Every absence answer includes label checked, date window, and jurisdiction source — models know what they’re seeing.
One join, done inside Kukun — your models receive a complete property record.
Integration
One API Call.
Nine Primitives.
This is what the join looks like in practice — the response shape your models actually receive.
Get /v1/property/details — response (abbreviated)
Ready to Stop Stitching?
API access, enterprise licensing, and MCP server configuration available now.
Talk to us about your use case.
For Homeowners
Remodel Cost Estimator
iHomeManager
Home remodeling guides