How a global bank turned a search engine into its data backbone

When BBVA's David Jiménez Ausin looks back at 2014, he sees a very different banking landscape. “Almost everything was still via web channel, as the [banking] app wasn't as developed as it is now, and each service had its information in its own systems,” he recalls. As strategy senior manager at BBVA, one of Spain's largest multinational banks serving 80 million customers worldwide, Jiménez Ausin found himself facing a common challenge: how to turn fragmented customer transaction data into one accessible system.

Working with Elastic version 0.9 in 2014–2015, BBVA's early trials showed promise, but when the team moved to testing with real users, they hit roadblocks. Performance lagged, some queries were painfully slow, and the system struggled to scale. “We were Beta testers at that time,” Jiménez Ausin explains. “We didn't know how many indices to have, or how many shards, replicas, sizes of each shard. It was very stressful.” But they persevered and by 2017, the new system could give customers access to 10 years of transaction history instead of just one. They could search their banking data by text, view complete aggregated information, and download their records — in milliseconds.

Today, the bank offers customers almost 20 years of searchable history. What began as 9 separate services has expanded to over 50. The system handles over 45 billion data points, yet maintains lightning-fast response times — from 5 milliseconds for simple queries to 200–400 milliseconds for complex financial health calculations.

BBVA is now exploring Elastic's frozen data tier to maintain its high service levels at lower cost while meeting compliance requirements. Frozen data tier, powered by searchable snapshots, allows financial services firms to store rarely accessed data in cost-effective object storage while maintaining searchability, reducing memory usage by loading data structures only when needed. It is fully integrated with autoscaling, allowing FSIs to seamlessly scale to store massive amounts of data

For Jiménez Ausin, BBVA’s experience holds a clear lesson: “Don't hesitate to at least try Elastic technology for financial projects. Give it a chance … a text search engine didn't seem designed for banking projects, but time has proven us right. It's a system capable of handling any transaction you can imagine at a tiny cost of development and maintenance, with extraordinary performance results.”