Often Health Care providers would like to track the disenrollment of members. This enables Health Care providers to tune their plans, provide best care and retain the members.
This workflow provides a close-to-real-time tracking of "recent patient touch impressions" via PowerBI.
As soon a recent patient touch is made by a care provider, data about the touch is loaded in the Lakehouse.
A scheduled data pipeline picks up the loaded information and runs through a prediction model. The prediction result is a prediction score indicating if a member will potentially disenroll in the future or not based on recent experience.
The prediction score is saved in a Lakehouse table.
The BI Report built on this table is updated. We can also set Alerts on BI Reports to send an email notification in case a threshold has reached.
The solution enables Primary Care Managers and decision makers with proactive and easy access to latest prediction scores.
Project name
Member Disenrollment Prediction
Description
Often Health Care providers would like to track the disenrollment of members. This enables Health Care providers to tune their plans, provide best care and retain the members.
This workflow provides a close-to-real-time tracking of "recent patient touch impressions" via PowerBI.
The solution enables Primary Care Managers and decision makers with proactive and easy access to latest prediction scores.
Project Repository URL
https://github.com/mikelenart/hack-together-fabric-ai
Project video
https://github.com/mikelenart/hack-together-fabric-ai/blob/main/video/Microsoft_fabric_hackathon_submission.mp4
Team members
abishtcca, igoyal01, mikelenart