Work we have delivered
Client names are withheld under confidentiality agreements. We are happy to walk through the technical detail on a call, and provide references at contracting stage.
SAP BusinessObjects to Power BI Migration
Client Challenge
A long-established BusinessObjects estate with hundreds of reports, unclear ownership and no reliable inventory of what was still in use. Licence renewal was approaching and the business had no view of what could safely be retired.
Solution
We inventoried usage before touching anything, retired reports with no consumers, and rebuilt the remainder on a governed Power BI semantic model rather than migrating report-for-report. Definitions were agreed once and reused across dashboards.
Technologies Used
Results Achieved
- Report estate reduced to what the business actually used
- Single agreed definition for shared KPIs
- Self-service enabled on certified datasets
Azure Data Factory Implementation
Client Challenge
Ingestion was spread across manual scripts and scheduled jobs with no central monitoring. Failures were typically discovered by the business rather than by the data team, and onboarding a new source took weeks of bespoke development.
Solution
We built a metadata-driven ingestion framework in Azure Data Factory, where each source is a configuration row rather than a new pipeline. Every run writes status, row counts and timestamps back to a control table, giving a single monitoring view across the estate.
Technologies Used
Results Achieved
- New sources onboarded by configuration, not new code
- Failures surfaced by alerting rather than by users
- Complete audit trail across every pipeline run
Microsoft Fabric Modernization
Client Challenge
Reporting ran on direct extracts from the ERP system, creating load on a production platform and offering no historical view. The organisation wanted a modern analytics platform without disrupting the operational system.
Solution
We staged hundreds of ERP tables into a Fabric lakehouse using a metadata-driven control table, layered into bronze, silver and gold, with quality checks and reconciliation at each stage. Reporting moved onto the gold layer with full history retained.
Technologies Used
Results Achieved
- Reporting load removed from the production ERP
- Full audited history retained in the lakehouse
- New tables onboarded through configuration
Data Warehouse Transformation
Client Challenge
A legacy on-premises warehouse with business logic buried in hundreds of stored procedures, no test coverage, and month-end runs that regularly overran into the working day.
Solution
We rewrote the transformation layer in PySpark against a medallion architecture, with reconciliation against the legacy system through a full parallel close cycle before decommissioning. Business logic was documented as it was migrated.
Technologies Used
Results Achieved
- Month-end processing completed within its window
- Business logic documented rather than tribal
- Parallel run signed off by report consumers before cutover
Want the detail behind any of these?
We can walk through the architecture, the decisions we would make differently, and what a comparable engagement would look like for your platform.