Customer Dashboard (APAC)
- SQL
- Power BI
- Databricks
- →No source data existed; built the dataset from the device up
- →21,000+ scanners and 14.5M+ scan submissions tracked from 2022
- →Became the base layer other dashboards were built on
Context
Most dashboards on the team started from a dataset that already existed in Databricks. This one did not. There was no consolidated source for scanner usage or scan submissions at all, and several teams wanted exactly that.
Reverse-engineering the data
To find out where scanner activity was actually recorded, I borrowed an iTero scanner and used its features one by one, then followed each action through the back-end databases to see what it wrote and where. Piecing those trails together gave me a usable dataset covering scans, feature usage, and Invisalign submissions.
The dashboard
On top of that I built a general-purpose Power BI dashboard, fed by SQL, that now tracks over 46,000 Invisalign providers, 21,000 installed scanners, 14.5 million scan submissions, and 6.6 million feature usages across 15 markets since 2022. It breaks submission trends down by year, quarter, and month; ranks the ten most-used scanner features split by on-device versus web access; and plots the distribution of how long doctors take to complete a scan, binned by the minute. Separate pages give a downloadable raw table for offline analysis and a pre versus post-purchase view of whether installing a scanner actually lifts clinic activity.
Result
It became a core reporting tool for commercial, marketing, and regional leadership, and the dataset underneath it was reused as the foundation for later dashboards, including the DSO performance dashboard.