The transformation of data platforms is being driven by breakthroughs in multicloud data management, intelligent applications and advanced data control planes.
These technologies are creating more complex and integrated systems, designed to manage enormous amounts of data while upholding stringent security and governance standards. With the growing demands of machine learning and artificial intelligence, there is a need for durable, scalable platforms. This technological integration is set to transform how businesses handle, analyze and utilize data, leading to more agile and sophisticated solutions.
“Our core business is data and analytics and information and insights company as such, data platform is at the core of anything that we do,” said Venkat Achanta (pictured), chief technology, data and analytics officer at TransUnion LLC. “We have differentially invested into our platform and [are] transforming our company … from a data and scores company into a more platform-centric company where we are leveraging the platform that we built. We are also offering platform-based products to our customers in this end-to-end data analytics value chain.”
Achanta spoke with theCUBE Research’s John Furrier and Dave Vellante at the Supercloud 7: Get Ready for the Next Data Platform event, during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. They discussed the advancements in multicloud data management, AI-driven applications and resilient data control systems, highlighting the shift towards more integrated and secure data platforms.
Building a next-generation data platform using multicloud data management
TransUnion, operating in over 30 countries and managing data for over a billion consumers, has transitioned from a data and scores company to a platform-centric organization. This shift is integral to handling the daily influx of tens of thousands of data sets and over 50 billion data points, according to Achanta.
“We curate that data, the data quality, cataloging of the data, more importantly connecting that data we call identity resolution, trying to figure out where this food this data belongs to and then serving up that data into intelligent predictive applications and cognitive applications for banks and the consumers to consume,” he said.
There has been the development of two key layers within TransUnion’s data platform, OneDev and OneTru, Achanta explained. OneDev serves as the infrastructure abstraction layer, enabling secure and efficient provisioning of infrastructure across AWS, GCP and private clouds. OneTru, on the other hand, focuses on the core business enablement aspect. It integrates and manages the vast amounts of data TransUnion handles, providing a foundation for developing machine learning models and delivering insights.
“It’s not just about cost for TransUnion, it’s [being] able to drive innovation faster. All of these silos, all of the different technology stacks is we’re not able to put the best foot forward from our innovation and time to innovate,” he said. “Cost is always a good thing if you can do it in a lower cost and a structured way that enables innovation. That’s the main model for TransUnion … we have converted our overall organization to let’s become platform centric and put a lot more of our weight into the core capabilities and the end-to-end data analytics value chain of … the core platform.”
Navigating the future with generative AI and data governance
As the data landscape continues to evolve, challenges and opportunities are presented by gen AI and data governance in the form of multicloud data management, according to Achanta. TransUnion’s approach involves a balance between using third-party products and developing proprietary models. The company employs a maker strategy for its core data platform, ensuring control over the inputs and outputs while leveraging external models for specific use cases.
“In our core data platform, we are going with a maker strategy. We don’t have to build all the way the LLMs and other things,” he said. “We are farther going down the automation chain for helping some of these users be able to accomplish leveraging gen AI.”
Governance and security are crucial, given the sensitive nature of the data TransUnion handles. The platform includes features for autotagging and managing data permissions, ensuring compliance with stringent regulatory requirements. Companies need to build a rock-solid infrastructure to support gen AI applications. There is a need for an integrated approach that combines structured and unstructured data to create comprehensive knowledge graphs, Achanta emphasized.
“To have a best gen AI infrastructure, you first need to start from what is your knowledge graph,” he said. “How do you simulate your structured data and unstructured data together? How are you going to be able to derive new knowledge through these gaps? These are all first-class concerns.”
Stay tuned for the complete video interview, part of SiliconANGLE’s and theCUBE Research’s coverage of the Supercloud 7: Get Ready for the Next Data Platform event.