![case study](https://cdn.quantiphi.com/2024/02/39b30419-analyzing-fraudulent-claims-in-insurance.png)
Analyzing Fraudulent Claims in Insurance
InsuranceBusiness Impacts
Fraud Claims Analytics
Billions of Dollars in Cost Savings
Customer Key Facts
- Location : North America
- Industry : Insurance
Problem Context
Fraudulent claims can cost companies billions of dollars per year. Building a dashboard on Looker can help insurance agencies improve the fraud detection process for future claims.
Challenges
- Obtaining claims data in .CSV format
- Pulling data into Google’s BigQuery and process the raw files
- Bringing data to Looker from BigQuery and visualizing multiple metrics
- Collecting labeled insurance claims data from open sources
- Determining key characteristics that lead to a fraudulent claim
- Creation of customized fields to generate relevant insights
Technologies Used
![Cloud Storage](https://cdn.quantiphi.com/2024/02/321aa23a-cloud-storage.png)
Cloud Storage
![Google's BigQuery](https://cdn.quantiphi.com/2024/02/da18ec97-bigquery.png)
Google's BigQuery
![Looker](https://cdn.quantiphi.com/2024/02/2348c9b8-looker.png)
Looker
Building a Looker Dashboard for Insurers to Detect Fraudulent Claims
Solution
The customer isĀ one of the largest provider of supplemental insurance in the United States, providing financial protection to more than 50 million people worldwide. They wanted to determine the probability of a claim being fraudulent based on historical data and understand the interrelationship between different dimensions in claims. Quantiphi built a dashboard on Looker that provides data visualization techniques and insight generation to help the customer better detect fraudulent claims.
Result
- Insights are derived from historical data, while also highlighting the characteristics of fraudulent claims
- Determines key relationships between features and helps create a solid foundation for machine learning
- Improved fraud prevention and detection process for future claims