Process and model large volumes of credit applications before the committed response time.
Identify hidden patterns and suspicious connections between applications.
Appropriately assess the risk probability.
Credit and insurance institutions face organized fraud through criminal rings that use false identities or hidden connections to obtain financing. During COVID, a company shared with us that it suspected an increase in organized fraud; however, the way its data was organized made it difficult to identify these suspicious relationships before credit approval.
Process and model large volumes of credit applications before the committed response time.
Identify hidden patterns and suspicious connections between applications.
Appropriately assess the risk probability.
Vinkos implemented a graph-based solution to analyze and visualize hidden connections between credit applications. A data ingestion process was designed to transform the applications into an optimized structure in Neo4j, enabling the modeling of relationships between data and the detection of suspicious patterns.
Thanks to this technology, the system was able to identify applications with a high likelihood of fraud, enabling the company to make informed decisions before granting financing. The data integration process ensured an efficient flow of information, optimizing analysis times.
With this solution, the institution achieved:
