DataVisor Introduces DCube
The Complete AI-Powered Fraud Management Solution
Scale is a term often encountered in technology space, but in the arena of fraud management, it takes on a singular resonance, as it serves to represent the staggering scope of fraud attacks permeating today’s digital economy. Fraudsters have a highly advanced set of tools and technologies at their disposal, and they are growing ever-more adept at using them to evade detection even as they’re ramping up the size and intensity of their attacks. This is what we mean when we speak of “fraud at scale”—waves of attacks, cleverly disguised, with the power to create widespread financial and reputational damage. Fortunately, we can also speak of “fraud management at scale.”
All here at DataVisor are very excited to announce the launch of DCube, our most comprehensive and advanced fraud management solution. DCube features a hyper-modern architecture that can manage complex digital signals at big data scale, and it enables all stakeholders within the fraud management ecosystem to collaborate on a single platform. The result is increased operational efficiency, significant financial savings, and frictionless customer experiences.
Real-time detection and high accuracy
DataVisor’s proprietary Unsupervised Machine Learning (UML) algorithms power DCube’s detection engine to go beyond simple anomaly detection, combining clustering techniques and graph analysis algorithms to discover correlated patterns that only the most advanced systems can reveal. This makes it possible to stop fraud in real time for even the largest enterprises in the world, and the resulting high detection accuracy and low rate of false positives mean good customers are never impacted. DCube additionally integrates aggregated intelligence from our Global Intelligence Network, along with supplemental supervised machine learning capabilities, further promoting high accuracy and low false positives—the latter being especially critical for preserving high-quality customer experiences.
From data to detection to prevention
Enhanced data management
Harnessing the full power of DCube begins with data validation and integration. These critical first steps help to establish a high quality of data early on; this, in turn, elevates model performance downstream, when models are deployed in production. For the user, the process is straightforward—upload data, and map user fields to DCube fields. From there, the system analyzes quality and flags potential issues automatically. As soon as data is validated and all data sources are integrated, users can engage in additional feature engineering for optimal model performance.
Advanced modeling capabilities
DCube offers a fully packaged set of features specifically designed and optimized for fraud and risk analytics purposes, and the inclusion of these exclusive features enables users to create high-performing models for any fraud scenario. The system will make informed feature and modeling recommendations to facilitate optimal performance, but users also have additional flexibility to customize input data, engineer complex features, and select models uniquely suited to the goals and challenges of their specific organizations. The overarching result is faster iterations, greater agility, and more time for data scientists to build high-performance models.
One of our primary goals in building DCube was to make it possible for large enterprises to accelerate adoption of advanced unsupervised machine learning capabilities while simultaneously enabling seamless collaboration between relevant stakeholders. By bringing together fraud experts and data scientists, and putting full control in their hands, DCube gives organizations the power to meet scale with scale, detecting even the largest and most sophisticated attacks in real time.
In offering a single platform where fraud and data science teams can seamlessly collaborate, DCube makes it possible for organizations to quickly align on strategy, efficiently evaluate detection results, and rapidly get new models into production.
Once models are built and deployed, fraud and data science teams can work together to review detection results, compare models, and improve performance. Then, with the insights gleaned from high-performing models, fraud teams can review cases, take bulk action on groups of correlated accounts, and set auto-decisioning rules to improve review speed by orders of magnitude. The ability to deconstruct events within fraud attacks, investigate complicated cases with detailed detection reason codes, and view correlated activities across all accounts, makes DCube an all-powerful tool for maximum fraud prevention, enabling organizations to adapt to fraud at lightning speed.
Our commitment to defeating fraud at scale
As a company, DataVisor is many things. We are domain experts in the field of fraud, and our expertise informs every feature of DCube. We are engineers, dedicated to building the most powerful tools and putting control into the hands of our clients. We are researchers, tracking every development across the fraud landscape to ensure we’re delivering precisely those capabilities organizations need to stay ahead of today’s fraudsters. We are a mission-driven organization that believes every business and every customer should be able to conduct personal and financial business online free of abuse, attack, and harassment. Above all else, we are a service provider in the truest sense of the term—we serve our clients by empowering them to protect themselves from fraud.
We are proud and excited to deliver DCube as the embodiment of our commitment to defeating fraud at scale.