Case StudiesCross-Ministry Common Client Identification
$34M - Annual cross-system cost quantified for the first time.
1,511 - Common clients identified across 3 ministries among 8,864 individuals - in three distinct population types.
Challenge
Three ministries suspected they were serving the same high-cost individuals. No shared identifier existed and privacy rules meant one could not simply be created. We found them anyway.
What we built
Probabilistic name and attribute matching across three ministry databases, validated to a measured accuracy, with hierarchical clustering to characterise the population. The method became the government's standing cross-ministry data sharing framework.
The model
Outcome
Method became the Government of Alberta's standing cross-ministry data sharing framework.
Client details and delivery specifics are confidential. References available on request.
