Power BI migration & social housing dashboards
The organisation
St Vincent de Paul Society Victoria is one of Victoria’s most prominent community service organisations, supporting some of the state’s most vulnerable people across a wide range of programs including soup vans, community housing, education and youth engagement, family violence support, alcohol and drug services, homelessness prevention, and over 100 retail shops across Victoria.
I joined the Reporting team within the Strategy, Missions & Innovation department, a team responsible for building organisation-wide reporting and business intelligence capability to improve decision making across all service areas.
The challenge
The organisation was in the middle of a significant data transformation moving from a fragmented Excel-based reporting environment to a centralised Power BI platform. With dozens of branches operating across the country, each storing data independently in Excel on the cloud, the data landscape was complex, inconsistent and difficult to consolidate.
The immediate goal was to build a monthly Board Operational Reporting dashboard that gave leadership a single, reliable view of organisational performance across all service areas. To get there, the data first needed to be found, understood, structured, and validated before a single dashboard could be built.
What I did
My work spanned four interconnected workstreams:
KPI identification and data needs analysis Working across all major service areas, I identified the appropriate indicators and metrics for monthly board-level operational reporting. This involved understanding what each area was actually measuring, what data existed, and how to bridge the gap between what leadership needed to see and what the data could actually support.
Data gathering and structuring I gathered raw datasets from operational areas across the organisation and structured them into a consistent format suitable for Power BI consumption. Given that source data was spread across multiple branches in inconsistent Excel formats, significant data wrangling was required to produce clean, reliable inputs. An engineer on the team then set up Azure pipelines to consume these structured Excel files into the Power BI environment.
Social housing dashboard One of the key dashboards I built was a tenant management dashboard for the community housing program. It provided operational visibility over current and new tenants, including demographic profiles, rent tracking, tenancy duration, and pipeline management. This replaced a manual, spreadsheet-based process that made it difficult to get a timely or accurate picture of housing operations.
KPI report refactoring I also supported the refactoring of existing KPI reports, improving their structure, accuracy, and usability for the reporting team and senior stakeholders.
The outcome
The end result was a meaningful shift in how the organisation accessed and used its data from scattered, manually maintained spreadsheets across disconnected branches to a consolidated Power BI environment that gave leadership a reliable, timely view of operations across all service areas.
The feedback from my manager at the conclusion of the internship captured the impact of the work:
“Your contributions to the dashboard builds and KPI report refactoring have been truly exceptional. These initiatives have provided tangible improvements to our reporting processes, which will have a lasting impact on our operations… you have not only solved some particularly complex problems but worked with a level of autonomy and initiative that is rare to find - even among seasoned professionals.” - Lucy, Manager, St Vincent de Paul Society Victoria
What I took from it
Enterprise data can be messy in reality and needs proper governance among fragmented sources, inconsistent formats and no centralised infrastructure. The side of business intelligence that people don’t see is the part where you try to understand the data well, behind the scenes of building dashboards.
The migration from Excel to Power BI also meant rebuilding how metrics were defined, not just where they lived. KPI mapping forced a level of precision about what the organisation was actually measuring and why, which turned out to be the harder and more interesting problem.