A petroleum company does not exist in isolation. It operates through licences, facilities, counterparties, contracts, storage arrangements, product movements and people. Yet most industry information systems still store those elements as separate records.
The result is a familiar problem: the information exists, but the relationships have to be reconstructed manually every time a decision is made. An entity graph changes that model. It treats businesses, licences, sites, depots and evidence as connected nodes whose relationships can be preserved and analysed over time.
Why relationships matter more than record count
A system can contain thousands of licence records and still provide very little intelligence. The licence number becomes more useful when it is connected to the correct legal entity, when the verification date is known, when historical changes are preserved and when the relevant facilities or trading relationships can be understood around it.
The same is true for storage. Capacity by itself is one fact. The relationship between a depot, its operator, the wholesaler using it, the period of access and the supporting evidence creates a much richer operational picture.
In petroleum intelligence, the important question is often not what records exist, but how the records relate.
The basic nodes in a petroleum graph
A practical petroleum entity graph can start with a limited set of high-value entity types:
- Businesses: legal entities, trading names and relevant organisational identities.
- Licences: petroleum licences and their status, scope, dates and verification history.
- People: authorised contacts, directors and other role-bearing individuals where lawfully relevant.
- Sites and facilities: retail sites, depots, storage infrastructure and other physical nodes.
- Relationships: supplier, customer, tenant, operator, storage, logistics or service relationships.
- Evidence: documents, verification records, test results and other supporting material.
- Signals: observations or events that may warrant review without being treated as established fact.
The value does not come from making the graph as large as possible. It comes from making the important relationships clear, governed and traceable.
Identity resolution is foundational
Before relationships can be trusted, the system needs confidence that two records actually refer to the same entity. Variations in company names, abbreviations, trading names and data-entry formats can easily create duplicates.
Identity resolution therefore becomes a core capability. Structured identifiers, verified registration information and historical aliases can help consolidate records without silently merging unrelated businesses.
Where confidence is insufficient, the system should retain the ambiguity rather than force a match. A probable connection is analytically useful, but it is not the same as a verified connection.
Time changes the graph
Relationships in the petroleum industry are not static. A wholesaler may move from one depot arrangement to another. A site may change suppliers. A licence may be amended. A director may leave. A storage contract may end.
A useful graph therefore needs temporal information. It should be possible to understand not only whether two entities are related now, but whether they were related at a relevant point in the past.
This historical dimension is especially important when reconstructing the context around a decision or event. Current data alone can create a misleading picture of what was true at the time.
Graph intelligence is not automatic guilt by association
Relationship analysis must be governed carefully. The fact that two entities share a facility, supplier or service provider does not imply wrongdoing or risk. Connections create context; they do not create conclusions.
This is why PetroleumOS separates relationship intelligence from judgement. A graph can surface a cluster, an unusual pattern or a connection worth reviewing. Any consequential conclusion should still be tied back to evidence, scope and a defensible analytical process.
Why this architecture matters for PetroleumOS
The PetroleumOS ecosystem naturally produces different relationship signals. LicenseCheck can contribute verification history. Storage Partners can contribute authorised infrastructure relationships. FuelGuard can contribute fuel-quality evidence. Report Fuel can contribute tightly controlled integrity signals. FWA and Petroleum Academy can contribute industry and knowledge context.
The entity graph provides the connective model across those specialist functions. It allows a verified signal created in one context to become useful elsewhere without losing its source or access restrictions.
From search to understanding
Traditional systems are designed to retrieve a record when a user already knows what to search for. Entity intelligence supports a different experience. A user can begin with a business and see the relevant licences, facilities and verified relationships around it. A new signal can be placed into its existing context. A changed relationship can trigger a review of affected records.
This is the shift from information retrieval to industry understanding.
PetroleumOS is being designed around that shift. The aim is not to map every possible connection. It is to preserve the relationships that materially improve trust, due diligence, operational visibility and strategic intelligence.