TDWI | Training & Research | Business Intelligence, Analytics, Big Data, Data Warehousing

Banking on Semantic Technology: AI-Powered Data Quality Balances Fraud Prevention and Customer Excellence

Deeper data connections protect compliance and optimize the customer experience in real time.

• By Phil Maitino
• January 17, 2020

Financial institutions are in the midst of a juggling act: working to reduce fraud, complying with a spectrum of Know Your Customer (KYC) and anti-money laundering (AML) regulations, and meeting their customers’ expectation for a smooth and convenient banking experience. At the heart of these operations is data quality, well-established technology that is today leaping ahead with the integration of artificial intelligence (AI) to create new advantages for regulated industries.

For Further Reading: 3 Ways B2B Companies Can Use Customer Data to Improve Customer Service in 2019 Modern Metadata Management Data Quality and the Single Customer View

Working together, data quality and AI-based tools such as machine reasoning are streamlining central operations that include identity verification and customer onboarding. With an improved ability to uncover deeper data relationships in real time, bankers are becoming more efficient and their systems are becoming smarter. It’s a powerful advantage that simultaneously reduces fraud and losses while sustaining excellence in customer relationships.

Managing Competing Priorities

Historically, it was common for different data formats, incomplete records, and nonstandard addresses to flow into an organization’s system, causing inaccuracies in database queries. This can result in customers receiving the same communication on multiple occasions, an unnecessary drain on the bank’s budget and a hit to customer service. Fraudsters -- for example, those requesting the issuance of multiple cards to the same person at different addresses -- can be tough to intercept when your location information is incomplete.

KYC and AML processes, although imperative to banking in the 21st century, can also expose organizations to fraud and regulatory risk. In addition, unwieldy processes related to KYC and AML can exacerbate dissatisfaction with customer onboarding; if the process seems extreme, invasive, or unnecessary, the customer is likely to walk away and may ultimately engage with a competitor who makes things simple. Keep in mind that customer expectations are high -- very high -- shaped by digital leaders (such as Apple and Amazon) that have inadvertently set the bar for any organization tapping into digital customer services and communications. These concerns were validated in a recent survey of fraud executives, with 88 percent of respondents expressing concern about the onboarding experience when choosing new tools for risk assessment.

Today it is possible to optimize both regulatory compliance and the customer experience, with trusted banking relationships powered by data and based on a real-time, 360-degree view of the customer. Simple yet sophisticated, integrated tools can validate, verify, correct, and enrich contact data as it enters the system to head off data quality challenges. With comprehensive global intelligence -- data quality coupled with AI -- bankers can decide instantly whether to accept new customers, detect application fraud in all types of customer channels, and ensure that only verified and standardized customer data enters the system, every time.

How It Works

A type of AI known as semantic technology, or semtech, associates words with meanings and recognizes relationships between them. Semtech is already proven in healthcare and pharmaceutical markets, helping streamline drug development with stronger data insights. It is this same value that works so well with the deluge of data handled by regulated industries. By applying context and making inferences from in-depth intelligence on bank customers, bankers and compliance officers create powerful, real-time connections among and between the spectrum of data sources they commonly use. ID verification is seamless, fast, and smart, addressing the needs of bankers and ensuring an ideal customer experience at the same time.

Machine reasoning and its ability to handle automated pattern recognition are built into semtech. This allows banking platforms to fill in any gaps in data -- for example, missing or incorrect information introduced in the application process. While these capabilities address core issues of data quality and completeness, they also help bankers better know their customers and which products and services may prove beneficial, extending the relationship.

These technologies also help reduce time spent correcting mistakes, ensuring properly validated identities using data that is curated, normalized, and integrated. From an operational standpoint, semtech can also facilitate regulated processes, such as building flexible, automated credit-checking and antifraud workflows. The competitive advantage is long-term because optimized global intelligence tools enable bankers to retire costly legacy compliance and KYC systems. This removes the need to spend time manually reviewing data and solving customer issues, freeing bankers to focus on product development and better customer service.

Data Is Everything

Semantic technology is only effective with accurate data. For example, when data quality is not part of an onboarding solution, the match technique between incoming identities and the repository relies only on the simplest form of exact matching. Without data quality, inferior identity verification engines cannot determine issues with critical fields, such as a missing street suffix, a misspelled street name, or an incorrect city name (North Logan for Logan, Utah). If these entries are standardized and corrected, bank systems benefit from more accurate matches that reduce the risk of untrusted IDs slipping through the cracks.

Semantically enabled machine reasoning offers an efficient use of AI that maintains basic requirements such as data quality and completeness, values that are scalable for ideal decision support in essential applications. AI (via semantic technologies) introduces a wealth of opportunity to increase efficiency in all matter of enterprise business applications. With this new and more extensive intelligence, errors are reduced, data insights are more quickly generated and sophisticated, and staff is empowered to focus on exceptional service, new product innovation, and overall business success.