OVERVIEW
Globally, workers compensation Insurance fraud has become a major concern for all the insurance companies which is continuously increasing year by year. The aim of this project was to build an AI model in order to rate the claims so that the SIU teams can process the cases according to the risk levels. This could lead to significant savings by eliminating fraudulent claims. This led to improved revenues and business continuity in the long term.
Key Benefits (Minsky)
User-Friendly, cloud-based AI platform
No coding skills are required for results or predictions.
Provides you a list of dependency features that can be used to optimize your business
Ability to fine-tune or optimize the models by trying different algorithms/prediction attributes
Easy integration with other third-party solutions such as TABLEAU for data visualization
Results
Greater accuracy in handling claims.
Identification of patterns in the data and fraudulent claims.
Reduction in manual effort to eliminate fraud.
Major cost reduction in fraudulent claims.
Can deliver more efficient fraud checks and audits.
Improved customer support
Executive Summary
The client is a large Insurance Company that provides Workers compensation Insurance policies to the workforce. The main challenge for them was to predict genuine vs. fraudulent claims. The company was not able to identify the fraudulent claims that were costing them billions of dollars in annual losses. The aim of this project was to rate the claims so that the Special Investigative Units (SIU) can process the cases according to the risk levels. After a detailed evaluation of their operation data, AI Labs (www.ailabsinc.com) used its proprietary Minsky AI Engine to build an optimized model using a combination of AI algorithms and prediction attributes. Based on this historical AI model, current customer claim data was used to predict if the claim could be fraudulent for each customer. This solution was developed and implemented in less than a week.
Typical Insurance Fraud Challenges:
How to minimize losses from fraudulent claims.
Adversely impacts the insurer’s relationship with its existing customers.
Has a significant impact on competitive advantage.
Excessive resource time wasted to check fraudulent claims manually.
Difficulty in identifying fraudulent vs. genuine claims.
Solution
After evaluating the client‘s challenges, we used Minsky to accurately model historical data of past customers claim data along with other related attributes for model creation. AI Models were generated by Minsky using historical data such as gender, age, occupation, nature of injury, location, injury description etc. for the selected Algorithm. Then predictions were generated for the current customers claim data to rank the claims from high risk to low risk for each claim. This allowed for proper resource allocation of the SIU teams for investigations according to the risk levels. Prediction data from Minsky was also integrated with 3rd Party data visualization application like Tableau. Robotic Process Automation (RPA) was then used to route the claims to the respective SIUs based on risk levels.