uplyst.Give enterprise data a shared business meaning.
The same customer, account or asset may look different in every system. Semantic context connects those records to definitions people can understand and use consistently.
Start with the definitionBusiness conceptOrder fulfillment
Customer commitment
The agreed obligation to a customer, recorded under the applicable commercial terms.
- Agreed delivery
- 24 Sep · 16:00 UTC
- Authority: approved agreement
- Current estimate
- 24 Sep · 18:00 UTC
- Authority: dispatch observation
Agree on what the data means
Customer, account and legal entity can mean different things in different applications. A semantic model records those distinctions: what each concept represents, its attributes and how it relates to the rest of the business.
Uplyst combines semantic modeling with enterprise information. This supports questions such as which customer commitments are exposed to a supply interruption or which assets support a critical service.
The meaning of “open” depends on the source.
- Receivables
- Open invoiceAn amount remains unsettled.
- Customer support
- Open caseA service request remains unresolved.
- Commercial account
- Active relationshipThe account meets its own agreed business rule.
Matching the word would lose the distinction. Model the fact, its domain and the rule behind it.
Keep identities and ownership clear
A shared concept can have several source identifiers. Retaining each identifier and its source helps distinguish a confirmed match from an uncertain association that needs review.
Ownership can differ by attribute. A CRM may own the account relationship while finance owns billing status. Conflicting values need an accountable owner; one convenient source cannot stand in for every system.
One entity, retained identities
Customer identities- Business entity
- Customer account
- CRM account
- crm:account:AC-104
- Finance customer
- erp:customer:CU-208
- Match evidence
- Reviewed legal entity reference
- Account owner
- Commercial operations
- Billing authority
- Finance operations
Check data against business rules
Required identifiers, allowed currency codes and order references are examples of constraints that can be checked. W3C SHACL defines a standard for expressing constraints over RDF graphs; other data models can use equivalent validation methods.
Validation alone does not establish accuracy. A correctly formatted order can still refer to the wrong customer, so structural checks work alongside reconciliation, business rules and review.
| Model element | Question it answers |
|---|---|
| Concept | What does this thing mean in the business? |
| Identifier | Which source record does it refer to? |
| Relationship | How is it connected to another concept? |
| Constraint | What conditions must be checked? |
| Owner | Who resolves changes or disagreement? |
Keep definitions current
A new business unit, contract type or source application can change how a concept is used. Versioned definitions and mappings make those changes visible to the teams and systems that depend on them.
Historical context matters too. If the definition of an active customer changes, an earlier report still needs to be understood under the rules used when it was produced.
Versioned business definitions
Definition v1
Active customer: an account with a completed order in the preceding twelve months.
Proposed definition v2
Active customer: an account with a completed order in the preceding twelve months or an approved live service agreement.
What requires review?
The population changes. The domain owner must review affected reports, effective dates and historical interpretation before the new rule is applied.
Resolve differences between systems
Different labels, conflicting owners and duplicate organizations are common sources of disagreement. A maintained semantic model makes those differences visible and records how each mapping was resolved.
Its value depends on how well people can apply and maintain the definitions. Coverage, traceability and the effort required to handle change matter more than the size of the vocabulary.
Build a shared understanding of your data.
- Use shared business definitions
- Retain source identities and mapping evidence
- Manage changes to definitions and relationships