Enterprise data architecture

Connected data.
Shared understanding.

We connect source systems, preserve history, and build shared models for reporting, managed self-service BI, and conversational BI. Proven architecture and automation give your teams a foundation they can use, govern, and extend.

Discuss your architecture

The big picture

From source data to useful answers.

Our reference design combines medallion layers with Data Vault integration, Kimball analytics, and shared semantics. We select and adapt these patterns around your business and platform.

  1. SourcesConnect systemsApplications, files & APIs
  2. BronzeCapture dataSource values & provenance
  3. SilverIntegrate historyData Vault keys & relationships
  4. GoldShape analyticsFacts & shared dimensions
  5. SemanticDefine meaningMeasures & access rules
  6. BI & AIExplore answersReports, Excel & data agents

Services across every layer

Shared standards. Connected delivery.

All six capabilities below support every layer: Bronze, Silver, Gold, and Semantic.

  1. Bronze
  2. Silver
  3. Gold
  4. Semantic

One foundation for delivery & operations

  • Lineage

    Catalog assets and trace their sources.

  • Orchestration & automation

    Coordinate repeatable data processing.

  • Quality & reconciliation

    Check completeness and reconcile totals.

  • Governance & security

    Apply ownership and access rules.

  • Monitoring

    Track pipeline health and data freshness.

  • Version control & CI/CD

    Review, test, and release changes.

Inside the architecture

A clear job for every layer.

Bronze / automated ingestion

Capture once.
Keep the source context.

Repeatable ingestion preserves source values and records where and when data arrived. Keeping capture separate from business rules gives the warehouse a traceable starting point for integration and reprocessing.

A dependable record of what arrived.
Source systems → preserved records
  • Business applications
  • Operational databases
  • Files & APIs
Bronze records Original source values Source identifier Load time & batch

Silver / Data Vault integration

Connect the business.
Retain its history.

Hubs hold business keys, links connect them, and satellites retain descriptive history. Our certified practitioners use Raw Vault to preserve source history and Business Vault to apply reusable rules as systems evolve.

Adaptable integration with an audit trail.
Illustrative Data Vault / customers & orders
Raw Vault · source history
HubCustomerBusiness key
LinkCustomer–orderRelationship
HubOrderBusiness key
SatelliteCustomer history
SatelliteOrder history
Business VaultReusable rules & derived business context

Gold / Kimball dimensional models

Make analysis consistent
across the business.

Facts describe business events at an agreed level of detail. Conformed dimensions give those events shared context, so each new information mart contributes to an enterprise model instead of an isolated report.

Comparable analysis across business processes.
Shared dimensions connect business processes
Customer
Product
Date
Sales factsOne row per sales lineQuantity · sales amount
Returns factsOne row per return lineQuantity · return amount

Semantic / shared business meaning

Define the logic once.
Reuse it in every report.

Semantic models turn analytical data into agreed measures, business names, and relationships. Central teams maintain definitions and platform-supported access rules while report authors and analysts build on the same governed models.

One consistent meaning across tools and teams.
From analytical structure to business meaning
Gold facts & dimensions

Shared semantic model

Measures
Revenue, order count, growth
Business context
Names, relationships, hierarchies
Access
Permissions for approved data

Managed self-service

Shared models. Independent exploration.

Central teams own the models and definitions. Business teams create reports, explore in Excel, or ask questions through configured data agents. Each team gets its own view without rebuilding core business logic.

Illustrative sales model · many ways to use it
Maintained centrally

Sales semantic model

Shared measures · relationships · access rules

Power BI

Build team reports

Connect thin reports to shared measures.

Excel

Explore the model

Analyze approved data in a familiar tool.

Conversational BI

Ask in business terms

Explore configured sources with a data agent.

“How did revenue change this quarter?”

How the architecture enables AI

Give the question business context.

Semantic models supply governed measures. Business vocabulary and optional ontologies add context about entities and relationships. We configure data agents around approved sources and evaluate answers against agreed business expectations.

An illustrative conversational BI workflow
Business question
“How has sales revenue changed by product category and quarter?”

Governed measures

Revenue definition, reporting periods, and approved semantic models.

Business context

Vocabulary and relationships; ontologies where they add value.

  1. InterpretMatch the question to relevant context.
  2. QueryUse configured sources and permitted access.
  3. ExplainReturn results in business terms.
Evaluate & refineCheck representative answers against shared measures and stakeholder expectations.

Designed for delivery

Evolve models and reports with confidence.

Models and reports have separate release processes. Validate each change and connect reports to the right model in development, staging, and production.

Two delivery tracks · three connected environments

Development

Shared modelsBuild & version
Team reportsCreate & iterate

Staging

Shared modelsValidate & review
Team reportsCheck connections

Production

Shared modelsPublish shared definitions
Team reportsRelease business views
Promote each track independently Reports use the model in their environment

Automate the standards.

Use repeatable pipelines, versioned SQL, or WhereScape templates. As a certified WhereScape partner, we tailor automation to your platform and team.

WhereScape expertise

Fit your platform.

Adapt storage, integration, and deployment to your environment. We bring considerable experience in SQL Server, Azure Synapse, and Microsoft Fabric, plus implementation experience in Databricks and Snowflake.

Platform expertise

Leave a foundation to extend.

Deliver in useful increments, with documentation, mentoring, and reusable patterns your team can carry forward.

How we deliver

Put the architecture to work

What should your next
data product enable?

Start with your business priorities. We will help shape the architecture and a practical delivery plan.

Let's talk architecture