Urban Decision OS · Reference Architecture
Municipal Intelligence Reference Architecture
A governed architecture showing how municipal sources become evidence, intelligence context, human-reviewed readiness, and executive control.
- Advisory only
- Human review required
- Evidence-linked
- No automated action
- Cloud-ready architecture
Architecture rail
- Sources
- Ingestion
- Warehouse
- Catalog
- Intelligence engines
- Decision readiness
- Control tower
- Sources
- Municipal, spatial, market, statistical, and governance references.
- Ingestion
- Batch, manual registry, future governed pipelines.
- Warehouse
- Governed analytical store · future cloud warehouse.
- Catalog
- Metadata, lineage, ownership, review status.
- Intelligence engines
- Rules, data science, macro / micro context, future AI gated.
- Decision readiness
- Readiness state, evidence sufficiency, confidence band, next human action.
- Control tower
- Executive view, reviewer queues, buyer-pack readiness.
Municipality capability map
A maturity ladder for municipal partners. Promotion through levels is governed by evidence, ownership, and review readiness — not by feature count.
Level 0
Evidence inventory
Enables
Catalog of candidate municipal sources and references.
Before promotion
Source families named; owners identified.
Governance caveat
No live integration claimed.
Level 1
Governed registry
Enables
Metadata, lineage, and review-status tracking per asset.
Before promotion
Metadata schema agreed; reviewer authority named.
Governance caveat
No automated approval.
Level 2
Readiness workflow
Enables
Reviewer routing, sufficiency gate, next-action prompts.
Before promotion
Review states defined; suppression rule in place.
Governance caveat
Human review remains the decision gate.
Level 3
Municipal review operations
Enables
Queues, senior-reviewer concurrence, audit trail.
Before promotion
Senior reviewer role appointed; audit policy approved.
Governance caveat
No regulatory determination implied.
Level 4
Intelligence engines
Enables
Rules, data-science, macro/micro context; AI is future-gated.
Before promotion
Engine inputs are catalog-traceable; governance rail in place.
Governance caveat
No automated decision authority.
Level 5
Executive control tower
Enables
Executive view of reviewer queues, readiness states, buyer-pack readiness.
Before promotion
Promotion from previous levels evidenced; control-tower scope signed.
Governance caveat
Advisory posture preserved at every level.
Where a cloud partner fits
Reference architecture only. Each surface below is a candidate deployment pathway that a cloud partner can support; none is a live or production integration today.
01
Secure ingestion
Can support governed ingestion patterns for municipal and reference sources.
02
Governed data platform
Identity, lineage, and ownership controls hosted on a governed platform.
03
Analytics warehouse
Candidate deployment pathway for the governed analytical store.
04
Metadata / catalog controls
Cloud-native catalog can support the registry's review-status fields.
05
AI / ML readiness layer
Future-gated wording / context layer above the deterministic brain.
06
Monitoring and access control
Reviewer identity, audit logging, and least-privilege access.
07
Partner deployment pathway
Candidate 60-day advisory pilot deployment pathway, not a production commitment.
UDI-OS owns the decision-readiness logic
01
Evidence posture
Source families, review states, and sufficiency gate.
02
Reviewer workflow
Queues, senior-reviewer concurrence, audit trail.
03
Readiness state model
Five states: needs review, metadata pending, reviewer cleared, gold set, suppressed.
04
Forbidden-claims guardrails
Linter-enforced banned wording across UI and copy.
05
Buyer-pack framing
Reviewer-cleared narrative bundle; no external export until clearance.
06
Human approval path
Senior-reviewer concurrence is the only decision gate.