Identity Management


Using artificial intelligence, a service company was able to find 80% of all fraudulent cases while only reviewing 40% of the total data

Artificial Intelligence is being used to help companies match identities across multiple data sources and draw important insights from them. A great use for artificial intelligence in identity management is fraud detection. Taxpayers have been known to claim “Homestead Tax Exemption” without having the necessary requirements. A reputable service company in Louisiana wanted to know if it was possible to reduce the amount of data needed to properly match identities across multiple data sources and utilize that data to detect Homestead Tax Exemption fraud.

They used a team that utilizes historical experiential rules/weighting system to match identities and determine the most probable cases of fraud. However, they were not happy with the data and time needed to properly detect fraudulent cases.

Where does AxonAI step in?

The first challenge was to find and correlate identities from multiple heterogeneous state-wide records, such as property tax, voter, and DMV data sources. Using our relational intelligence engine, we were able to find and stitch together multiple identities, across multiple data sources, to form a single connected identity for each tax exemption applicant. The next challenge was to find accurate homestead addresses. Using deep multi-variate analysis across those multiple databases, we were able to find and correlate applicant’s different addresses utilizing the characteristics of each record to determine tax exemption validity. Our goal is to reduce the need to review 80% of the cases in order to find the top 80% of fraudulent cases .

How did we do?

Using Axon:Verify, we utilized ten million identity records from different databases in Louisiana to build the taxpayer identity profiles. After we built those profiles, we were able to find 80% of the fraud while requiring review of only 40% of the cases, allowing for significantly improved use of the analyst to focus on the biggest ROI. We were also able to achieve this result without the need to purchase expensive background data. This gives the customer the flexibility to accept and process cases more efficiently, or reduce the amount of human analysts needed to process those cases. We achieved these results by connecting, at a very granular level, the specific characteristics of each identity record that linked the most to Homestead Tax Exemption fraud. With Axon:Verify, customers can make sure that identity data they own is utilized by extracting actionable intelligence in the most efficient and accurate way possible, even when a direct match does not appear feasible.

Need help matching identities across various databases? How about drawing advanced insights from that data?