Entity Resolution – how we spot the fincrime in financial data others miss
Every day, banks and financial services generate an enormous amount of data – transactions, and customer information, to name a few dimensions. With the volume consistently growing, the connections between different data points become increasingly fuzzy. This makes it harder to get valuable insights out of financial (and other) data and turn those insights into actions. This is of particular relevance in the fight against financial crime and money laundering where it’s more important than ever to quickly use available data and make accurate compliance decisions. Existing systems can’t effectively handle the increase in data, modern criminal behavior, and innovative financial services that change how regular customers store and use their finances. This is the need that drives Entity Resolution for financial institutions.
What is Entity Resolution?
Entity Resolution combines all data points into a meaningful and trustworthy single view that enables financial institutions to tackle compliance challenges in a data-driven, and therefore accurate, way. It tells you precisely whether multiple records in financial data match a real-life person or company, making sense out of names, bank accounts, addresses, e-mails, phone numbers, etc. It does this even when an overwhelming amount of inconsistent data is involved. This single view, created by Entity Resolution, detects unusual behavior and alerts you when criminals try to obscure their identity and unlawful intents by leaving out important information or providing it incorrectly.
Examples of these cases in Money Laundering and Fraud include:
- A company’s structure is conspicuous (e.g. Someone is the owner of several companies)
- A company provides no actual service but only shifts money back and forth
- Someone uses a layering scheme to hide the money’s origin
- Someone uses the services of more than just one bank or financial institution
- Someone consistently uses different identities with the help of disposable e-mail addresses or burner phones
- Bots trying to take on human identity by using apparently correct information as an alibi
Entity Resolution is generally dynamic and applicable in various use cases. This allows it to detect such cases by finding correlations that humans and legacy systems cannot recognize in high-volume datasets, since they only become apparent in the bigger picture. For example, e-mail addresses can look legitimate at first glance, but in a broader context show a continued similarity in names and their connection to specific bank accounts.
Hawk AI and Entity Resolution
Hawk AI uses Entity Resolution as one method to power its AI-enhanced products such as Transaction Monitoring and Payment Screening. But we don’t stop leveraging its strengths there; we apply it in real-time, combining internal and external data sources, and augmenting it with an additional tech stack such as entity graphs and Elastic Search. This makes Hawk AI’s Entity Resolution exceedingly fast, accurate, scalable, ultimately transparent, and data secure.
Entity Resolution in real-time
Applying Entity Resolution in real time ensures the foundation for data-driven compliance decisioning is up-to-date. This allows AML compliance and Fraud prevention teams to react faster and more dynamically to apparent threats or odd behavior. This is especially crucial in fraud scenarios since irreversible financial damage can occur if criminal activities are not detected quickly enough.
Greater coverage through internal and external data sources
Hawk AI combines internal and external data sources when looking for odd behavior in financial data using Entity Resolution. This means that information from a given dataset (e.g. Transaction data from one single organization) and external data (e.g. Data from trade registers) is used to create a comprehensive view. An isolated look at someone’s transaction data might not fully justify suspicion, but balancing it against trade registers or other external sources provides a fuller picture.
Entity Graphs and Elastic Search
It makes sense to visualize things to understand the technology behind Entity Resolution better. While Entity Resolution is the technological process of finding matches in the data, an entity graph is an underlying foundation where all the data, and connection points between data, is stored. If certain matches in the entity graph link to one specific individual (entity), a new subcategory within the entity graph is created: a dossier. A dossier is much like an individual file (imagine the medical file that doctors keep) that exists for a person or company. There, all the data points connected to this person or company like address, e-mail, or phone number, are graphically highlighted.
Hawk AI found entity graphs with their dossiers represented in a graph database inefficient for financial crime detection since it involves a lot of manual searching for the correct dossier. Instead, Hawk AI uses Elastic Search to navigate within the Entity Graph. There, the data is represented according to the created dossiers and is far better structured for financial crime use cases. Like a search engine, Elastic Search automatically finds all the information about the dossiers needed for Entity Resolution.
Flexible deployment of Hawk AI’s entity resolution – AML and Fraud use-cases
North American Bancard – Suspicious merchants
Hawk AI’s customers use the power of entity resolution to solve various AML and Fraud (FRAML) challenges.
North American Bancard provides hundreds of thousands of businesses with an end-to-end infrastructure for globally preferred payment types and processes more than $45 billion in electronic payments annually. These high customer and transaction numbers need to be monitored efficiently with the right technology to prevent cases of financial crime. NAB, therefore, uses Hawk AI’s transaction monitoring to:
- Flag merchants who were banned from NAB’s service for illegal activities try to sign up with a new, modified identity
- Find out if merchants own more than just one business and if these businesses are financially connected
- Detect merchants using their service for money laundering by being the merchant and the end-customer at the same time, or by using fake end-customer networks
German online retailer – Bot-networks as end-customer
In the use-case of a publicly traded German online retailer, Hawk AI’s transaction monitoring and entity resolution detect fraudulent behavior and money laundering schemes of end customers on its marketplace. It not only identifies suspicious activity by humans but also from internet bots. For example, Hawk AI’s technology discovered a bot network that tried to disguise itself behind the identity of apparently real people. The bots then ran a money laundering scheme where they bought products on the company’s marketplace just to return them to the merchants for a refund. With the help of entity resolution, Hawk AI’s technology could identify the connection between the various bots and detect the high amount of money distribution in connection with an unusually high product return rate.
Fintech-Unicorn – End-customer layering schemes
A South American fintech unicorn providing cross-border payments tackles fraudulent account activity and money laundering layering schemes with Hawk AI’s entity resolution. The technology identifies conspicuous bot-human transactions and, more specifically, illegal pass-through activity. Here, criminals use transitory accounts linked to themselves to transfer money from one account to another and back to themselves, trying to obscure the money’s origin. Entity resolution matches the accounts to real-life counterparts by connecting the various data points in the layering scheme. It picks up the circular money flow, which is not noticeable to the human eye.
The fight against financial crime needs modern tools like Entity Resolution
Criminals become more sophisticated in their fraud and money laundering schemes every day. They find creative new ways to avoid existing fraud and AML compliance systems, are technologically savvy, and often know how to leverage modern technology faster than financial institutions. At the same time, the data financial institutions need to monitor to guarantee regulatory compliance becomes increasingly multi-layered and complex. With Entity Resolution, Hawk AI offers organizations state-of-the-art technology to come out on top in the fight against financial crime.