Knowledge Vault

The city's source of truth—documents, streams, and systems unified into a governed knowledge graph with citations and lineage.

What It Is

Knowledge Vault is LaplaceX's canonical store for city entities, events, and documents. It ingests from parking, mobility, safety, energy and back-office systems, normalizes schemas, resolves duplicates, tags sensitive fields, and preserves evidence trails so every answer is explainable.

Why It Matters

Single Source of Truth

One unified graph for assets, places, incidents, and policies—eliminating data silos and conflicting information across departments.

Explainability

Every fact links to citations and version history—complete data lineage ensures decisions are traceable and auditable.

Governance by Default

Policy tags, data residency, and access scopes are enforced automatically—compliance is built into the architecture.

How It Works

1

Ingest

Connectors for APIs, streams, files

2

Normalize

Schema registry & data contracts

3

Resolve

Entity matching & merge logic

4

Tag

PII, purpose, region classification

5

Link

Build graph relationships

6

Serve

APIs for CQ, Insights, Orchestrator

7

Audit

Lineage & version tracking

Key Capabilities

Connectors: SQL/APIs/streams/IoT + document/CSV/PDF ingestion

Schema registry & data contracts (versioned)

Entity resolution (rules + ML) & conflict policies

Policy tags: PII/PHI, residency (UK/EU/MENA), purpose limits

Full lineage: source → transformation → consumer

Time-travel/versioned snapshots

Vector & graph indexes (semantic + relational)

Citations panel for every answer (used by Cognitive Query)

Export to BI/SIEM; SDK & REST/Graph APIs

Entity & Lineage Explorer

Explore how city data becomes a unified, traceable knowledge graph

Data Sources:
Parking APIWork OrdersIncidents CSV
Knowledge Graph
locationOfoccursOncreatesCarPark 101Incident #42RoadSegment AWorkOrder 77
Assets
Events
Documents
PII Data
Click a node to inspect lineage & citations
Entity Details
Selected Entity
CarPark 101
Data Lineage
Source: Parking API @ 13:40 UTC
Transform: normalize(schema=v2)
Transform: entity_resolve(cp101)
Policy: standard access
Citations & Sources
s3://city/parking/2025-09-14/daily.json
api.parking.local/v1/carparks/101
Connected Entities
Incident #42(locationOf)

Proven Outcomes

↑ Accuracy
Query accuracy + explainability ↑; blind spots ↓
Zero Duplicates
Duplicate records merged; fewer handoffs
Minutes
Audit prep time: minutes, not days

Use Cases

Parking & Mobility

Canonical car-park and occupancy history for accurate predictions. Link parking events to traffic patterns and maintenance schedules.

Learn More

Public Safety

Incident–asset–work-order linkage for faster triage. Connect emergency responses to infrastructure and historical patterns.

Learn More

Energy & Facilities

Asset registry with complete maintenance lineage. Track energy consumption patterns and predict infrastructure needs.

Explore Platform

API Integration

Programmatic access to the knowledge graph with full lineage and citations

Entity Resolution
// POST /api/kv/resolve
{
  "type": "carpark",
  "candidates": ["cp101", "cp101_alt"],
  "policy": "prefer:official_id"
}

// Response
{
  "id": "cp101",
  "merged": ["cp101_alt"],
  "lineage": [
    {
      "source": "parking_api",
      "timestamp": "2024-01-15T14:30:00Z",
      "confidence": 0.95
    }
  ]
}
Entity & Lineage
// GET /api/kv/entity/cp101?with=lineage,citations

// Response
{
  "id": "cp101",
  "attrs": {
    "name": "CarPark 101",
    "capacity": 450,
    "location": [25.12, 55.17]
  },
  "edges": [
    {"to": "inc42", "label": "locationOf"}
  ],
  "lineage": [...],
  "citations": [
    "s3://city/parking/2025-09-14/...",
    "api.parking.local/v1/carparks/101"
  ]
}

Frequently Asked Questions

How do you prevent hallucinations?

All responses are grounded in the Knowledge Vault. If there's no evidence for a claim, the system returns no answer rather than guessing. Every fact includes citations to source data.

Can we keep data in-region?

Yes—data residency tags automatically route storage and compute to appropriate regions (UK, EU, MENA, etc.). Processing stays within designated boundaries for compliance.

How is access controlled?

RBAC/ABAC with row and column-level filters ensure users only see authorized data. All access is logged with full audit trails for compliance and security reviews.

Turn Scattered Datasets into an Explainable Knowledge Graph

Unify your city's data sources with complete lineage, governance, and explainability. Transform fragmented information into a trusted foundation for AI-powered decision making.