One of semtech’s primary values is that it can update and deliver new data and learnings as soon as additional data is input into the system. This is a groundbreaking ability, allowing the enterprise to redefine concepts and relationships as new information becomes available. It means semtech has a long-term role to play in the enterprise, creating impact as a flexible and powerful tool that evolves in step with data.
Semantic layer improves current and long-term data value
Combining machine reasoning, machine learning and traditional data quality tools makes it possible to deliver a flexible, interoperable semantic integration layer on top of enterprise systems.
Figure 1: Combining traditional data quality methods with effective AI delivers the most effective data quality and master data management environment available.
This layer makes the application of both main forms of AI (machine reasoning and machine learning) approachable and efficient. The supports improved data quality and interoperability, the ability to use these technologies to identify and align data to different standards, and finally, to apply advanced AI-based pattern recognition applications out of the box.
Application benefits of semantic machine reasoning
With data quality and interoperability issues addressed, AI opens the opportunity for advanced data applications in all types of enterprise settings.
Advanced banking applications
By identifying and surfacing patterns, semtech helps in quickly identifying possible customer risks. For example, in a proofing process which includes national ID and age verification, semtech can automatically flag suspicious individuals who appear on any of dozens of Office of Foreign Assets Control and European Union watchlists. Not only does this minimize risk, it empowers smarter decisions on what to do next: approve, deny, or escalate. Further business advantage comes from the ability to integrate data from virtually any source. This means semtech can spot trends, such as new account growth or an increase in fraud in a certain geographic area. The business is empowered with the potential to predict possible future growth or similar trends.
Product data quality
Product data quality can get messy quickly, for example in the event of a merger or acquisition where monumental amounts and types of data must ultimately work together. This scenario is often characterized by different platforms and protocols for data storage—harmonizing all this information for tangible business value is a significant challenge. AI-enabled data quality lets data managers build out their product descriptors or lexicons. These are used to infer relationships among and between products, allowing similar or complementary products to be better identified and managed. With semtech’s inherent data interoperability, data quality is seamless and presents the business with a new world of insight based on new data relationships.
Streamlining customer onboarding
Financial institutions must comply with Know Your Customer (KYC), anti-money laundering (AML), and similar financing regulations. At the same time, they face the challenge of delivering on the customer’s expectation for convenience, speed, and simplicity. In addition, they must work diligently to mitigate the risk of fraud. While fraud detection is important, customer expectations demand a frictionless experience. Hurdles that are perceived as extreme, invasive or unnecessary can cause immediate abandonment. AI-enabled data quality, linking, and pattern-detection technologies can extend data’s reach to help meet regulatory and practical goals that identify forms of identity fraud with reduced false positives.
It’s all about the data
It is critical to note that semtech is only effective with accurate data. For example, without data quality tools woven into onboarding operations, the match technique between incoming identities and the repository is inherently inferior, relying only on the simplest form of exact matching. Less sophisticated identity verification engines are unable to determine issues with critical fields, such as a missing street suffix, a misspelled street name, or a city name (i.e., North Logan for Logan, Utah). By standardizing and correcting these types of issues, the enterprise benefits from more accurate matches and reduces the risk of untrusted IDs slipping into its master data management system.
Semantically enabled machine reasoning is an efficient form of AI that can help with basic tenets like data quality and completeness, and that can scale to provide automated pattern recognition for decision support in mission-critical applications. AI—delivered by semantic technologies – opens a wealth of opportunity to improve efficiency in all types of enterprise business applications. In this very powerful new era, errors are reduced, data insights are more sophisticated and quickly gleaned, and staff is freed to focus on excellent service, new product development, and overall business growth.