Crash-analysis dashboard

The problem

Internal teams and application vendors needed to understand how applications behaved on real devices in the field, not in a test lab. Without device-level visibility, crash investigations were slow and largely anecdotal.

The data existed but lived on a legacy Oracle platform behind manual refreshes, which made it unreliable and a blocker to decommissioning that platform.

What I did

I built and published an application analytics dashboard covering roughly 174 applications at daily install volumes between 290 thousand and 1.3 million, plus a campaign dashboard reporting on millions of eligible devices across 18 campaigns.

I migrated the source off legacy Oracle onto Snowflake and replaced manual refreshes with scheduled production jobs, removing a blocker to decommissioning the legacy platform.

A related analysis across roughly 75,000 users over six weeks surfaced an 11% decline in daily application launches, which fed into product and activation decisions.

The outcome

The dashboard enabled device-level crash analysis for both internal teams and vendors, and supported a crash-reduction initiative the business sized between $500K and $1M.

It also retired a manual process and unblocked a platform decommissioning, the kind of second-order win that never shows up in a screenshot.