AUGUR
{{ authUserEmail }}
{{ authTitle }}
{{ authMessage }}
{{ authError }}
Don't have an account? Sign up Already have an account? Sign in
⚠
Couldn't update watchlist
{{ watchlistError }}
Geospatial Intelligence Atlas

{{ regionTitle }}

{{ positiveCount }} candidate parcels flagged across a {{ totalCount }}-parcel {{ regionSummaryNoun }}. Zoom and pan the live map, or click any amber marker for its full readout.
⚠
Live predictions are forward-looking — outcomes are not yet known.
Generated {{ liveGeneratedDate }}. Ranked by a model fit on historical outcomes, scored against today's parcels — not a record of what actually happened.
Spatial Context
Real parcel coordinates from the {{ regionSubLabel }} {{ regionContextNoun }}.
Live Rendered
{{ lg.label }}
Parcel Registry
{{ positiveCount }} positive signals
#{{ row.rank }}
{{ row.apn }}
{{ row.topSignal }}
{{ row.coordLabel }} {{ row.scoreLabel }} Positive ›
Geospatial Intelligence Atlas

District Intelligence

A synthesis of regional spatial analytics, predicting rezoning through multi-modal intelligence vectors.

{{ d.tag }}
{{ d.name }}
{{ d.sub }}
{{ d.growth }}
Top-Decile Lift
BASE RATE
{{ d.rezoning }}
Positives
{{ d.inv }}
Validation
{{ d.risk }}

Intelligence Streams

{{ s.n }} {{ s.title }}
{{ s.text }}
← Back to District Assets
In-Depth Backtest Report

{{ reportMeta.label }}

{{ reportMeta.reportSub }}
{{ reportMeta.statusVerdict }}
{{ st.label }}
{{ st.value }}
{{ reportMeta.liftNote }}
Signal Weights (documented, not fitted — no label exists to fit against)
{{ w.label }} {{ w.w }}
Methodology Notes

{{ reportMeta.narrative }}

{{ reportMeta.caveat }}
No interactive parcel map is published for this jurisdiction yet.
Cross-Market Validation Program
Six-Market Validation Sweep
4 OF 6 MARKETS ROBUST DATA VINTAGE: JUN–JUL 2026
Validation Status IN PROGRESS
Robust Markets
4/6
Maricopa, Phoenix, Charlotte, and Raleigh clear the robustness bar. Nashville is thin-but-real; Franklin has a rule-based shortlist, not a backtest.
{{ m.label }}
{{ m.value }}
{{ m.sub }}
{{ chartTitle }}
{{ chartSubtitle }}
{{ b.value }}
{{ b.label }}
validation_pipeline_log.txt
{{ l.text }}

Simulation Archive

The full multi-T validation history behind every headline number on this site. District Assets and the Report pages show one T=2022 snapshot per market — this is every cutoff actually tested, so nothing here is cherry-picked from a single good year.

Observation Window
2018 – 2024 (5 markets swept)
Runs Logged
21 backtests + 1 signal build
Validation Method
Spatial GroupKFold, 8km cells
{{ mk.label }}
{{ mk.sub }}
{{ mk.statusVerdict }}
T-CutoffPositivesBase RateLiftROC-AUC
{{ run.T }}{{ run.positives }}{{ run.baseRate }}{{ run.lift }}{{ run.roc }}
No backtest sweep — Franklin has no historical old-zone/new-zone field to validate against, so this is a rule-based signal shortlist, not a fitted model.
{{ mk.footnote }}
Data Governance

Compliance Intelligence

Every source this platform reads from is a public government record — parcel rolls, assessor sites, zoning-case trackers. No scraped listings, no private data, no PII. Non-disclosure states are excluded from coverage rather than silently degraded.

{{ r.label }}
{{ r.value }} {{ r.sub }}
Market Coverage Map
6 Live Jurisdictions — AZ, NC, TN
12 non-disclosure states excluded from coverage (sale price is not public record there). See per-jurisdiction status below.
Jurisdiction Data Status
{{ row.name }}
{{ row.sub }}
{{ row.framework }} {{ row.status }} {{ row.sync }}
Intelligence Feed
Automated Verification Logs
{{ f.time }} — {{ f.tag }}
{{ f.text }}
Saved Parcels

My Watchlist

Parcels you've starred from any market's backtest or live shortlist. Removed automatically only when you remove them — not when rankings refresh.

Sign in to save parcels

Star any parcel from a market's Parcel Registry to track it here across sessions and devices.
Nothing saved yet

Open any market's Spatial Map, find a flagged parcel in the Parcel Registry list, and click its star to add it here.
Saved Parcels {{ watchlistCount }} saved
{{ row.apn }}
{{ row.topSignal }}
{{ row.regionLabel }} {{ row.modeLabel }} {{ row.scoreLabel }}
Documentation

{{ docsContent.title }}

{{ b.h }}

{{ b.p }}
The AI Rezoning Model

The AI that predicts rezoning
before it happens.

Backtested across six markets of real historical outcomes — not promised, proven.

Optimal Zoning Alpha simulation
Strongest Result: Charlotte, NC
8.95× Top-Decile Lift, T=2022
Validated Lift
Point-in-time features, spatial GroupKFold backtests — repeatable and leakage-controlled, not live-updating.
Top-Decile Lift, Best Market
9.5×
Kernel Output
class RezoningBacktest:
def score(parcels_at_T):
features = build_pit_features(parcels_at_T)
// spatial GroupKFold, 8km grid cells
probs = cross_val_predict(model, features)
return rank_by_lift(probs, base_rate)
# Output: Charlotte, T=2022 -> 8.95x top-decile lift
1.43M
Parcels Screened
1,002
Verified Rezonings
Obsidian monolith
The Inference Engine

The receipts behind every score — not marketing copy.

Six markets. One validation standard.
Every score on this platform traces back to a documented backtest or an explicit rule-based rubric — real historical outcomes, real base rates, real confidence intervals.
AUGUR
Predictive Intelligence
© 2026 AUGUR Predictive Intelligence. All rights reserved.