30+ Legacy Reports Rebuilt on a Modern BI Stack, with One Definition per Measure
Static reports differed between hospital facilities and the same measure meant different things in different places. We rebuilt more than thirty reports, reconciled every definition, and automated delivery.
- Service
- AI Business Applications
- Industry
- Revenue Cycle & Finance
- Region
- United States
- Engagement
- Lab-type ODC (dedicated team)
Outcomes
- 30+ reports modernised
- One definition per measure
- Hands-free delivery
Stack
Power BI · Power BI Report Builder · SQL · SharePoint
Written byOriental Nippon Team
A US revenue-cycle operator produces reporting for a large number of hospital client facilities. Over the years those reports had been built on a legacy reporting platform and on an enterprise application's built-in reporting.
The result was thirty-odd static reports in which the same measure could mean different things, and a facility that wanted something slightly different got a new copy of a report instead of a parameter.
The problem
- Static output. Facilities could read the reports but not explore them.
- Conflicting definitions. The same measure was calculated differently in different reports.
- Manual production. Building and distributing the set was a recurring job for the team.
- Copies instead of parameters. Facility-specific needs were handled by duplicating reports.
| Before | After | |
|---|---|---|
| Reports | 30+ static, legacy platform | 30+ interactive, Power BI |
| Measure definitions | Varied by report | One documented definition per measure |
| Facility variations | Separate copies | Parameters on one report |
| Production | Manual, recurring | Scheduled refresh, export and delivery |
What we built
30+ rebuilt reports
Financial, batch performance, stair-step recovery, invoice summary, detailed invoices, month-end, ACH, aging and acknowledgement reporting, each tailored to the facilities that use it.
One definition per measure
A semantic model and metric dictionary, so a number means the same thing everywhere.
Automated delivery
Scheduled refresh and export, emailed and placed in SharePoint on time.
A decision log
Every definition conflict and how it was resolved.
Technology and the role it played
| Technology | Role in this project |
|---|---|
| Power BI Desktop | Interactive reports on one semantic model |
| Power BI Report Builder | Paginated, print-ready report formats |
| SQL | Source queries behind each measure |
| SharePoint and scheduled email | Automated delivery to facilities |
How we delivered it
Most of the real work was not in Power BI. It was in getting agreement on what each number means, and writing those decisions down.
- Inventory. Every report, who reads it, and what question it actually answers.
- Reconcile definitions. One agreed definition per measure, with the decisions logged.
- Rebuild in waves. Grouped by consumer, so each facility's set changes once.
- Parallel run. Old and new compared for a full cycle before anything was retired.
- Automate. Production and delivery switched on once the content was agreed.
01Inventory
2 wks
Every report, consumer and measure catalogued
02Definition reconciliation
2–3 wks
One agreed definition per measure, decisions logged
03Rebuild waves
8–12 wks
Reports rebuilt in consumer groups
04Parallel run
2–4 wks
Old and new compared for a full cycle
05Automation
2–3 wks
Scheduled refresh, export, email and document-store delivery
Risks and how we handled them
| Risk | How we handled it |
|---|---|
| Conflicting definitions rebuilt into new reports | Definitions reconciled and logged before rebuilding |
| Facilities facing constant change | Rebuilt in consumer groups so each facility's set changed once |
| New reports that disagree with old ones | A full cycle of parallel running before retirement |
Outcome
Facilities now receive interactive, standardised reports tailored to their own needs, produced and delivered without manual work.
The bigger change is less visible. Two facilities comparing the same number are now comparing the same thing.
What the client keeps
- Report definitions
- The semantic model and metric dictionary
- Refresh and distribution configuration
- The reconciliation decision log
Sources and further reading
- What is Power BI?, Microsoft Learn
- Paginated reports in Power BI Report Builder, Microsoft Learn
- Conformed dimensions, Kimball Group
- The dbt Semantic Layer, dbt Labs
Facing something similar? This is the kind of work our AI, Data Science & Automation team delivers every day. Talk to us about your stack.
