Hear how TransUnion handles scaling data-heavy transactions for fraud prevention and risk solutions using Aerospike and Adaptive Map.

In this session, Ramesh Kumar Manickam, Senior Director of Technology, TransUnion Global Capability Center, Mohanraj Vanjiappan, Senior Manager, TransUnion Global Capability Center, and Bharath Gowda, Architect, TransUnion Global Capability Center, discuss challenges faced during data migration and highlight their technical solutions.

Highlights covered:

🌱 TransUnion’s digital authentication solutions authenticate consumers behind devices for various transactions and industries, flagging high-risk interactions for digital fraud prevention while reducing friction for customers.

🌱 TransUnion’s data-heavy ecosystem utilizes Aerospike cluster with all-flash configuration, storing billions of records containing identity and transactional data, updating nearly 4 billion records daily.

🌱 Challenges during data migration included splitting time series data across multiple records due to Aerospike’s record size restriction, maintaining custom ordering, and ensuring sub-millisecond to millisecond response times for transactions.

🌱 TransUnion implemented Adaptive Map, a customized solution, to dynamically scale and split records, enabling efficient data management without code changes. They incorporated Expressions, a domain-specific language, for server-side business logic, enhancing performance and scalability.

🌱 Batch operations and asynchronous calls were utilized to optimize data loading and reading, achieving response times within milliseconds for customer AP transactions, and loading 60k+ transactions per second.

-🌱 Custom charts provided by Aerospike allowed TransUnion to monitor batch operations effectively, ensuring 99% of batch operations were completed within 90 milliseconds and 99.9% within 150 milliseconds.

🌱 TransUnion successfully handled global customer transactions, demonstrating the scalability and efficiency of their adapted solution.