Copy of Case Studies.png

About The Client.

The Requirement.

The Client’s current system of collecting health care (physicians, claim and miscellaneous) data from different sources and using Models to score physicians was running into performance issues as result of the volume and a flat, de-normalized data model.. This caused operational productivity issues and the need to move to a more modern, streamlined and normalized data source.

Solutions We Offered.

  • Cluster Build-out – AWS cluster build out with Spark

  • Data Model Transformation – Transformed the data model from de-normalized Vertica tables to normalized Impala tables

  • Data Pipeline Development – Developed initial data pipeline in Spark to populate normalized model

The Result.

The client was able to observe significant time reduction in transforming data from sources to the Hadoop cluster and also attained the ability to query data in real time while training the models before scoring.