INSIGHT / Case Study

Delivered a twin-module analytics solution for our client from the insurance sector.

Case Study - Analytics, Big Data, Insurance

Promise Delivered.

About Client:

Our client is a leading health insurance provider in the north-west (WA), catering to about 2 million customers with a network comprising about 40,000 doctors, hospitals, and healthcare providers.

Client Objective:

Make healthcare work better for all its customers.


Client wanted to get a 360° view of its customers to be able to help it attain the stated objective. In order to do that, they needed to analyze customer service calls, customer surveys, campaign reviews, emails, etc. Along with that, they wanted to have real-time ADT information from the various providers that the client serves. Client planned to build a data-lake in Hadoop that can hold such variety of data, along with providing a safe and secure environment for storage and analysis.

MSRCosmos delivered a twin-module solution to meet specific objectives.

Module 1: Objective:
Client received calls from customers and providers on a daily basis. Client wanted to analyze what the customer is calling about, what are the most frequent requests, how to cluster them, etc.


The call information contains audio files along with some metadata, which are not enough to undertake effective analysis. Therefore, the audio files were sent to an external vendor to mine the calls so that they can provide transcripts and scores, and additional metadata associated with those.

Upon the same being processed, MSRCosmos’ solution pulled the data (from the vendor) and stored it to Hadoop. This was followed by transformations on top of that data and load it into a set of parquet tables. These tables were later used by data scientists who would apply machine learning algorithms on top of the data.

Solutions Architecture for Healthcar Business


Module 2: Objective

Collect the data from various providers’ EMR systems in real-time, so that client can take action in advance and provide better quality of care to customer.

This was critical, as earlier the client had to submit a lot of paperwork to the providers to get any information which entailed a lot expenditure in terms of money and time.

Client approached the providers to share the required data in real-time (for which they’d be incentivized). That meant, although client had to pay for the providers willing to provide the data in real-time, them doing so had a very significant impact on the company’s objectives, as getting ADT data in real-time could help them take immediate action where required. This would also significantly reduce paper-work, thus helping the client serve its customers well and retain them.

However, since security is very important for the data hosted in Hadoop, the solution had a Kerberized cluster along with fine grained access to different Hive tables using Sentry. It also had encryption at-rest and encryption in-transit for all the respective data. That besides, the solution also comprised usage of TLS that would facilitate highly secure external communication of the client systems.

Contact us today discuss how we can help your business benefit from digital transformation. Or write in to us with your queries – our team shall respond to you with the relevant options and assist you in using the suitable solution.
Email: or T: +1 925 399 4218.


Contact MSRCosmos

Contact Us