Data expert Amy Hodler examines how one of the world’s largest financial services companies has successfully employed graph technology to fight fraud and better understand customer behaviour
Fraud is a growing concern in the financial services sector as criminals use increasingly sophisticated techniques that make it difficult to identify fraudulent activity at scale, and take fast, decisive action.
As financial fraud’s scope and size grow to include digital and mobile platforms, it is increasingly challenging for financial services institutions to stay ahead of criminals. As a result, fraud prevention and detection are at the top of the boardroom agenda. The problem is traditional methods of monitoring fraud are not strong enough to fight fraud. This is because fraudulent patterns are becoming increasingly complex.
The technique of setting up rules to examine deviations from standard purchasing patterns, using discrete data, for example, may work in catching lone wolves. It is not, however, an adequate tool to hunt down fraud rings. These organised groups use synthetic or manufactured identities, which are often stolen. These malevolent actors are also capable of hijacking mobile devices unknown to their owners.
Complicated fraud schemes demand an advanced way of following the chain that links one account to another. The secret is in spotting how activities, which at first may appear totally unrelated, are actually connected. This necessitates a 360-degree view of the intricate network fraud rings are made up of. It allows financial institutions to pinpoint suspicious events that may be linked with a high degree of accuracy.
Graph database technology may be the ultimate weapon in the war against financial fraud.
Fraud detection architecture using graph technology
Graph databases are capable of uncovering relationship patterns that are difficult to detect with traditional relational databases. For this reason, a growing number of the world’s leading financial institutions are adopting graph technology. Graph technology is being used to model and monitor data about customers, accounts, devices, locations and so on in a bid to identify fraudulent activity and take proactive action. Allianz, a multinational financial services company offering insurance products and services to 100 million customers in more than 70 countries, is one such company.
The company takes a zero-tolerance stance on fraud, which is a problem both in underwriting and when a customer makes a false claim on their policy.
Relational data model headaches
Allianz recognised the best way to understand any possible fraudulent activity is by storing, analysing and visualising claims cases through connected data.
“Graph technology does this at scale, which means we no longer have to rely only on highly demanding, traditional relational technologies,” explains Dr. Jan Doumen, strategic lead for Customer & Broker Information and Insights at Allianz.
The company had historically found it challenging to build internal visualisations of suspicious behaviours using relational technology. The approach was also not able to equip the Allianz Benelux team with useful data as and when it was needed.
By contrast, graph technology recognises potentially fraudulent activity in Allianz Benelux’s ecosystem by revealing concealed illicit connections. Correlating customer data into a graph database also enables the Allianz Benelux anti-fraud team to expose the risk exposures in a motor vehicle or household contents context, for example.
Positive business benefits
It is just as important for the Allianz Benelux team to have a 360-degree view of the customer. Its operations have been through a series of mergers and acquisitions recently. As a result, its customer data has ended up in silos, which has triggered operational inefficiencies.
“When we were able to get to a level with graphs to show colleagues this holistic view of a customer, it was so much easier for them to understand rather than through a table with rows and columns. This will enable them to personalise their services towards our customers,” adds Doumen.
Allianz has already seen real business benefits from its native graph approach. Over a two year period, a staggering €2 million of operational profit value was identified. Following the success the Allianz Benelux team has had utilising graph technology, it is now planning to roll out the technology to other areas of the organisation.
Unfortunately, fraud rings are only going to get cleverer. Graph databases have the power to future-proof an organisation’s fraud prevention initiatives by providing valuable data insight based on data relationships and connected intelligence. A highly powerful weapon in the war against financial fraud that financial institutions cannot afford to ignore.
The author is Director, Analytics and AI Program at Neo4j, the world’s leading graph database company. She is co-author of Graph Algorithms: Practical Examples in Apache Spark & Neo4j, published by O’Reilly Media