Finding the source of a metric takes detective work
A report depends on several tables, files and transformations that are not documented.
Data Architecture
Every important number should have a traceable path back to its source
Architecture is not a diagram for its own sake. It should explain how data moves from a source system into an indicator, a report, an exchange or an API.
When this becomes a problem
Data Architecture
Integrations are growing independently
Every new exchange introduces its own formats, identifiers and business rules.
BI is fixing source system problems
Dashboards spend their time cleaning and stitching data instead of using a prepared analytical model.
What we work on
What the work covers
I look at architecture from both ends. How the systems work today and how leadership expects to use the result. The target design has to connect those two realities.
- Inventory data sources, tables, integrations and information flows.
- Define conceptual, logical and analytical data models.
- Design Data Warehouse, data marts and a semantic layer for reporting.
- Trace data lineage from source fields to indicators, reports and APIs.
- Design target integration patterns, master and reference data and change controls.
How I work01 02 03 04
How I work
Inventory
Map sources, exchanges, tables, reports and dependencies.
Model
Separate business entities from their current technical implementation.
Target design
Design integration, storage, marts and BI layers.
Transition
Sequence change without breaking the processes the organisation still relies on.