Data Analytics and Business Intelligence Dashboards

A set of Power BI dashboards that give divisional and central leadership a live view of operational performance. The dashboards are fed by Power Query transformations over the operational database, with SQL doing the aggregation where the logic needs it. The result is a monitoring surface that reflects current operational state rather than a monthly snapshot assembled after the fact.

Operational monitoring across the divisions was based on figures that were compiled and circulated after the period they described. Leadership could see what had happened, but not what was happening. Because the underlying data lived in operational systems shaped for transaction capture rather than for analysis, producing a view of current performance meant rebuilding the logic each time.

What made it hard

• Separated the dashboard layer from the operational data layer — dashboards must not become a second source of truth, only a second view of the first.
• Centralised the metric definitions in the transformation layer so that a KPI means the same thing on every dashboard.
• Used Power Query for the reshaping that would otherwise be repeated per report, and SQL where the aggregation was better expressed as a query.
• Built dashboards around operational questions rather than around available data fields.

A two-stage pipeline feeding a Power BI presentation layer. Power Query applies a shared set of transformations that turn raw operational tables into analysis-ready shapes, with SQL handling heavier aggregation. Because the transformations are shared, a metric is defined once and reused across every dashboard, which removes the inconsistency that comes from each report rebuilding its own logic. Power BI presents the result for operational monitoring across multiple divisions.

Source

Operational data held in the MySQL database — the same system of record the applications write to.

Transformation

Power Query applies shared reshaping and cleansing, with SQL handling aggregation where a query is the right expression.

Metric definition

KPI logic defined once in the transformation layer and reused across every dashboard.

Presentation

Power BI dashboards built around operational questions for divisional and central leadership.

• Reviewed the existing reports and identified which questions leadership actually needed answered, and which figures were being recomputed from scratch.
• Built the Power Query transformation layer to produce consistent, analysis-ready datasets.
• Moved aggregation into SQL where the logic was complex enough to be clearer as a query.
• Designed each dashboard around a specific operational question rather than around available fields.
• Validated the dashboards against the manually-produced figures before they became the reporting view.

• Power BI dashboards for operational monitoring across multiple divisions.
• Shared Power Query transformation layer defining metrics once.
• SQL aggregation for complex logic.
• Current-state monitoring rather than post-period snapshots.
• Validated against manually-produced figures before rollout.

• Gave divisional and central leadership a live view of operational performance instead of a delayed summary.
• Removed repeated reconstruction of the same figures across reports by defining each metric once.
• Scaled monitoring across divisions without a proportional increase in manual reporting effort.

Screenshots for this project have not been published yet.