PPavionExecutive Cockpit

Ontology & Data Mesh

The logical layer that lets a half-integrated roll-up still answer one question consistently — model once, federate the data, generate insights anyway.

Pavion · FY26 (modeled)
#4 SDM Top Systems Integrators (2025)
2,800 employees · 70+ US sites · 23 countries
🧩 Roll-up & integrationStep 1 of 7 · the mesh & data products behind itCompany HierarchyAll journeys
🌐 Enterprise 360 modules· on Ontology & MeshBrowse all 31 views ▾
● LiveBuilt forCIO / Data· integrate logically, not physicallyCFO / FP&A· one number across many ledgersIntegration PMO· insight before full migration

Pavion can't wait for every acquired ERP to migrate before it gets answers. The fix isn't one warehouse — it's a shared ontology (so everyone means the same thing) over a data mesh (each brand owns its data as a product), with a semantic layer that federates them. Insights generate today; they just carry a confidence flag where a brand isn't integrated.

Data backing: enterprise ontology (Doc 03) · knowledge graph (Doc 04) · semantic layer (Doc 05) · brand · site · org
Shared meaning (T-Box)

The enterprise ontology — what the words mean

Ten classes everything maps to. The Office is the keystone: it's where brand, leader, entity and geography reconcile.

Platform
Company1
Pavion (parent platform)
operates ▾ / owns ▾
The 'who' — accountability & ownership
Business Unit3
Fire · Security · Integration
Brand / Legal Entity10
acquired operating companies
Leader (Person)16
org / accountability
operates ▾ (brand → office)
The keystone
Office37
the reconciliation point
located in / serves / delivers ▾
The 'what & where' — delivery & demand
Region6
geographic rollup
Customer10+
accounts served
Contract
PX / ON-X / monitoring
Device412k
monitored installed base
Supplier6
equipment partners
Relationships (predicates)
Pavion operates Business UnitPavion owns Brand / Legal EntityBrand rolls up to Business UnitBrand operates OfficeLeader accountable for Business Unit / BrandOffice located in RegionOffice serves CustomerCustomer holds ContractContract monitors DeviceOffice delivers Job / ProjectSupplier supplies Office / Job
Federate, don't centralize

Each brand is a data product on the mesh

86% of revenue is already office-grain actual; the rest is read in place from legacy systems and reconciled — no big-bang migration required.

Pavion (core)
Integration · domain data product
Actuals
data quality / grain88%
AFA Protective Systems
Fire Safety · domain data product
Actuals
data quality / grain98%
Firecom
Fire Safety · domain data product
Actuals
data quality / grain100%
RFI Enterprises
Security · domain data product
Allocated
data quality / grain82%
Star Asset Security / Ion247
Security · domain data product
Allocated
data quality / grain72%
The Protection Bureau
Security · domain data product
Actuals
data quality / grain92%
ECD Systems
Integration · domain data product
Allocated
data quality / grain80%
(ISC)
Security · domain data product
Region-only
data quality / grain45%
Signet
Security · domain data product
Region-only
data quality / grain45%
DavEd Fire Systems
Fire Safety · domain data product
Actuals
data quality / grain95%
10 domain data products (above)
Federated semantic layer
entity resolution · canonical metrics · grain tags
Consumers
Story · Briefing · 360s · Simulator
Defined once, computed everywhere

Governed metrics — the logical layer

Every metric has one definition and a grain. The layer federates it across integrated and legacy domains, flagging where a value is allocated.

MetricDefinitionGrainHow it federates across brands
RevenueΣ recognized revenueoffice · jobactuals where integrated; allocated from area where not
Adjusted EBITDArevenue − COGS − SG&A (+ add-backs)BU · entityentity P&L normalized to one chart of accounts
ARRrecurring contract valuecontractfrom monitoring/PX systems across all brands
Recurring mixARR ÷ revenueBUfederated — same formula, many sources
DSOAR ÷ revenue × 365entity · officelegacy entities measured at area grain, flagged
Gross margin(revenue − COGS) ÷ revenuejob · BUmapped via canonical cost categories
Net retentionexpansion − churn on basecustomerresolved across duplicate customer records
The payoff

How insights generate before integration finishes

1 · Resolve

AI matches legacy entity / brand / office codes to one canonical node — so RFI's data lines up with everything else.

2 · Federate

Query reads each brand's data product in place; the semantic layer maps native fields to canonical metrics.

3 · Allocate + flag

Where a brand reports at area level, AI disaggregates to office on learned drivers and marks it an estimate with a confidence band.

4 · Reconcile

Allocated parts must tie back to the source total; anomalies and duplicate customers/suppliers across brands are surfaced.

This is not theoretical — it's how this cockpit already works. The Story, Briefing and 360 views read the same governed metrics over integrated and non-integrated brands alike; 86% of the numbers are office-grain actuals and the balance is AI-allocated and labelled. As each brand integrates, its data product's grain rises and estimates flip to actuals — the mesh closes itself.