Faster, More Accurate Search Saves Lives at UVA Health System
UDig’s Innovation Studio built a multi-system index allowing for unified search experiences. Over time this solution will provide weighted results to further improve and optimize the system’s accuracy, thereby decreasing the time spent searching for answers. In perspective, the time saved when a nurse needs to quickly find the dosage of a specific medication could be the difference between life and death.
STRATEGIC SNAPSHOT
Challenge
Replace aging and out of support Google Search Appliance that hindered medical professionals during critical times of need.
Strategy
Build multi-system index allowing for unified search experiences.
Outcome
A fully functional, robust and customizable solution which provides weighted results to further improve and optimize the system’s accuracy, thereby decreasing the time spent searching for answers.
The solution has had a dramatic impact on patient care; the time saved when a nurse needs to quickly find the dosage of a specific medication could be the difference between life and death.
Challenge
UVA Health System engaged UDig to replace an aging and out of support Google Search Appliance. The existing solution, while adequate, never truly provided search results that allowed users to react quickly while in the midst of a medical crisis. The goal, as articulated by the Director of Health Systems IT, was to create a search system that could literally save lives by providing faster search results to medical staff.
Outcome
UDig’s Innovation Studio team partnered with the client to assess solutions and quickly landed on SOLR. SOLR, an open source project based on Lucene index, is a very popular, highly scalable and supported platform with community support / plugins. UDig worked closely with the client to spin up AWS instances mirroring their environment in our cloud space, delivering a fully functional, robust and customizable solution in just a few short weeks. The build combined results from 2 different CMS systems (including a Plone instance and SQL Server databases) housing important links users could now access as a unified result while searching on any of their front end systems.
How We Did It
Tech Stack
- SOLR
- AWS
- Lucene
- Plone
- SQL Server
