Published:
October 10, 2020
Categories:
AI, Machine Learning, Natural Language Processing, Data Analytics, Healthcare, Insurance
Client:
POLMED

Project Brief

W e were given a sample set of three years (2017 – 2019) worth of transactional data for a particular medical insurance scheme. Though the dataset was in some parts unable to paint a complete picture of the history of the scheme transactions for a number of reasons not limited to, but including records which were missing relevant data in some fields for the aforementioned period, it was sufficient for extrapolating trends and patterns over the course, and providing a contextually comprehensive report of the nature of the transactions and the current business implications thereof. We
also automated the analysis and reporting to streamline the process for future data analysis requirements.

Project Outcome

We structured this report to efficiently outline two focus areas for each province; purchase trends per ATC5 code and purchase trends per Practice type. This report formed part of a digitization journey we are taking with the client to provide them with the ability to deliver an accurate account of where the business currently stands in real-time as opposed to relying on actuarial teams which take weeks and a countless number of hours to compile such reports.