Using Big-Data to Help an Insurance Company Identify Frauds in Insurance Claims to Prevent Losses
A renowned insurance company was facing a serious challenge to identify legitimate claims that were costing them serious revenue loss. To help prevent further damage to their business because of false claims, the company approached Glorious Insight for help and assistance.
Upon an extensive analysis, they found out that the company was using traditional data management infrastructure. This resulted in making the work of fraud detection complex since it took hours to process the data. Check out how Glorious Insight used multiple open source search technologies to resolve the problem.
Customer Background
The insurance company had grown significantly ever since its inception over the years but they were not able to process the customer’s data in a smart manner to identify frauds. Collecting terabytes of data, analyzing them, initiating the pay-ins & pay-outs in the form of premiums and claims was difficult. The difficulty was further aggravated with botched or unethical claims. Though they had an in-house fraud detection team for fraud protection, in the absence of an agile and scalable data analytics solutions, losses almost seemed inevitable.
To overcome this problem, they collaborated with Glorious Insight to set-up a data management system that could analyze numerous claims and evaluate them to make it fit for processing the funding support.
Challenge
There are multiple queries that are raised on a daily basis for insurance claims, settlements, and premium payments. As a result of that, huge chunks of data get collected over time and it becomes tough to analyze the pattern for fraud prevention. When the insurance company wanted to rummage through the data for insights to mitigate the frauds via predictive analysis, Vis-a-Viz understanding the pattern of the claims for most frauds. They choked out because of;
- The enterprise technology was not scaled up as per the changing technological requirements. They relied on a system that was too redundant to process a multitude of data, making efforts laborious as new data requests had to be sent to analysts who used to run SQL queries to compile reports. It was a time-killing approach to get to actionable insights.
- As traditional methods or infrastructure was used, a simple calculation that could have taken minutes consumed hours or even days for result generation.
- A lot of time was wasted doing manual computation of bill scans or scribbling the record numbers manually on data servers and extracting the same from the server individually. That was too much hassle to handle for agile operations.
Solution
After analyzing their old legacy systems, Glorious Insight suggested search technologies that could help build an agile, intuitive modern analytical platform on open source platforms.
- A customized platform was created for the purpose of content processing and indexing.
- To streamline faster access of data from lakes and warehouses, Glorious Insight developed a customized search solution to allow greater access to data achieves.
- Setting up a Query Processing Language for aggregation of queries with real-time charts, graphs, and results.
- A customized graphical presentation was designed for user-friendliness and intuitive presentation.
After deployment of these solutions, there was significantly less time-consumption since the anti-fraud investigator desktop dramatically enhanced scalability by processing data in significant volume with precision. The team that was designated to formulate SQL queries were relieved since data was available in real-time. Multi-faceted search further improvised the experience of data drillers to retrieve individual scans that forged better legal suit against scams.
Result
Customized search, predictive analysis, and quality assessment solutions helped flatten the database. On account of deployment of search customized big data interface;
- The enterprise could easily analyze over 20 million + claims with over 100 million details of bills processed at a lightning-fast speed. Such insights helped identify red-flags through suspicious activities to prevent monetary damage.
- Now the enterprise was able to import a billion scans and compile evidence so that proper legal suits can be filed with a sufficient amount of proof.
- To present the cases, the pieces of evidence were delivered in minutes instead of days for faster judgment and prosecution.
