Our Fortune 100 healthcare client was struggling to detect and quantify the damages caused by fraudulent provider activities. They needed to identify providers that are part of the network and are directly or indirectly involved in fraud.
- Detect the amount of damages caused from fraudulent provider activities
- Identify providers that are part of the network and are directly or indirectly involved in fraudulent activities
By working with the HyperAspect A.I. Cloud platform, our client developed an analytics application that analyzed internal and external data in order to flag and evaluate malicious activities by in-network providers.
The system continuously monitors the websites of the US Department of Justice and the Office of the inspector general to extract any postings related to active healthcare fraud verdicts. Once this data was internalized via stream-based pipelines and generation of crawling/scraping technology from the website, advanced artificial intelligence algorithms extracted key parameters from the unstructured texts. This data was first cross-referenced against NEPPES data and client’s internal provider network data. All of this complex data acquisition, augmentation and analysis was then visualized on a dedicated dashboard that allowed our client to observe records of malicious activities within their provider network in near real-time fashion.
Key requirements for Hyperaspect were to:
- Collect external publicly available unstructured data then augment it with governmental and internal enterprises data and apply an array of top of the line NLP, ML and deterministic approaches in a continuous manner
- Visualize the data in a business-friendly way in order to enable the fraud department to act and prevent further damages
After a month of operating under the newly created system, our client:
- Identified 612 questionable providers potentially involved in fraudulent activities
- Reduced damages against the national and private insurance funds by 12%
- Cleaned up 500+ malicious providers from the network
- Improved providers’ data reliability by 250%