In an recent interview with SafetyDetectives, Serpil Hall, the Head of Fraud Prevention at Celebrus, delved deep into the innovative strategies that set Celebrus apart in the world of digital fraud prevention. With over two decades of experience in combating financial crime across various industries, Hall brings a seasoned perspective to tackling both internal and external fraud. She highlights Celebrus’ unique approach to building and utilizing detailed digital identity profiles that begin tracking from the very first user interaction, a practice not common in the industry. Hall’s insights reveal how Celebrus’ real-time data processing and sophisticated behavioral analytics preemptively halt fraudulent activities, setting new standards in the data analytics landscape.
Can you tell us a bit about yourself and your role at Celebrus?
My name is Serpil Hall, and I am the Head of Fraud Prevention at Celebrus. I am a proven financial crime and fraud management professional with over 20 years of commercial experience across many industries and sectors with deep expertise in fraud operations, fraud prevention systems, fraud biometrics, and internal fraud controls. I have worked with some of the world’s largest global banks, airlines, and merchants on strategic and tactical improvements to reduce both internal and external fraud.
How does Celebrus approach fraud prevention, and what makes your strategies unique compared to other data platforms?
Celebrus as a platform is deployed in 32 different countries today and our mission is to improve the relationship between brands and consumers via better data.
Our Fraud applications all start with digital identity and our ability to build, persist, and provide detailed digital identity and profile data for both anonymous (pre-login) and known (post-login) consumers while also stitching this data together across all channels and devices used by those consumers instantly. Evidence should start from the first interaction, not just after a login. This is a common gap in the industry, and we are seeing incredible value where our customers use these evidence profiles to stop the fraud before it happens.
Celebrus captures, contextualizes, enriches digital data, and delivers (generates alerts/signals) it to downstream systems within milliseconds. Celebrus tracks interactions, behavioral biometrics, and journeys of individuals in real-time to detect and prevent fraud. Using a set of pre-defined and user defined scorecards, Celebrus detects anomalous/fraudulent behaviors and generates fraud alerts/events in real-time to prevent fraud.
Celebrus dynamically raises fraud alerts based on the behaviors exhibited by the individuals and automatically structures and collates the alerts in the standard Celebrus Data Model delivered to a data warehouse of the customer’s choice. Backed by the out of the box Celebrus Data Model, they can then use and make sense of digital fraud data better and quicker than any other systems can do.
Celebrus can also intervene to drive real-time prevention to make changes to the user experience based on the evidence and our identity and profile capabilities. For example, Celebrus, in real-time, identifies high application fluency such as the repeated use of copy-and-paste in the application form, along with low familiarity of personal information such as editing of name and address, and together they signal fraud, and an alert is raised which means our platform stops the transaction. The alert, at the same time, can also be sent to any other fraud management system pending further investigation.
Can you explain the importance of real-time data processing and how Celebrus manages this at scale?
Celebrus is deployed for our customers in a single tenant, private cloud with no shared infrastructure across our customers. Our system makes data available in milliseconds, in a fully processed manner, which means that brands can act quickly. Speed, and better data, are the two most important factors to staying out in front of the fight against fraud in digital.
What are the biggest challenges facing the digital data analytics industry today?
- Digital Identity: Most organizations today still work with flawed identity solutions that provide broken identities and disjointed profiles. If you want to fight fraud, you need to focus on ways to connect the dots of a single consumer’s activity (even if they are anonymous visitors).
It’s crucial to recognize the identity of the entire customer and their journey – across channels, devices, and interactions. That’s the only way you can truly know your customer and identify evolving threats before they become a serious problem.
- Profile Management: Organizations still struggle with user profiling, identity verification, anomaly detection, risk scoring and more. It’s important to apply profile management methods to detect and prevent fraud, protect sensitive data and safeguard the trust of the customer.
- Data Quality: Poor data quality, including inaccuracies, inconsistencies, and incompleteness, can undermine the effectiveness of data analytics initiatives. This is even more debilitating in the fraud world.
Where do you see the future of data analytics heading in the next five years?
- AI and Machine Learning: As AI technologies become more accessible and make it easier to use by anyone, this will empower organizations to extract actionable data insights from large and complex datasets more efficiently and quickly. However, the data quality challenges today will inhibit this progress if brands don’t rise up to that challenge and build better digital datasets.
- Real-time Data Analytics: There is massive demand for real-time data analytics capabilities. Technologies such as quantum computing and edge computing will facilitate the processing of data in real-time, allowing organizations to gain insights and act on them in the moment.
- Data Visualization: Data visualization tools and techniques will continue to become more immersive and interactive, and it will aid in many ways to explore and communicate insights with others. Augmented reality (AR) and virtual reality (VR) technologies may be increasingly utilized to create immersive data experiences, enabling users to interact with data in three-dimensional spaces.
- Data Analytics (compliant with laws and regulations): Organizations will need to prioritize transparency, accountability, and fairness in their data gathering and handling processes to ensure that data analytics initiatives respect individual rights and complaint with local laws and values.
- Data Democratization: Explainable data will continue to grow bigger. For example, ChatGPT and other open-source tools will carry on empowering non-technical users with self-service access to data and analytics tools. User-friendly interfaces, empowered by natural language processing (NLP), and augmented analytics capabilities will make it easier for users to get insights from data without relying on heavy lifting of data scientists or IT specialists.