A global financial institution uses SQream DB for fraud detection

Download customer success story: Global financial institution unlocks new insights


Finance and Banking

Millions of customers

A multinational credit card and payment services provider has to maintain trust with its customers

Consumer trust is paramount. To protect customers and retailers against fraud, the company’sData Science team needs to conduct long-term historical exploration and analytics on tremendous amounts of data.

The problem - Long running queries and limited analytic scope

The company's Data Science team was experiencing several challenges in the data preparation and feature generation stage of their AI pipeline on theshared Hadoop infrastructure they were using. Long-running queries were frequently interrupted or de-prioritized by other system users, and some complicated window function queries could not run at all.

The requirement

A solution was required that would alleviate these challenges, while supporting the company’s focus on maintaining their customers’ trust with advanced AI fraud detection techniques.

Waiting for query

The solution - Long term analysis for improved fraud detection and prevention

After testing several solutions, the Data Science team integrated SQream DB into their Spark and Hadoop-based architecture to power accelerated mass-data analytics.
The new architecture offloads important data to a SQream DB analytics server, enabling the Data Science team to run advanced ad-hoc queries, prepare data, and generate a large number of features to feed the AI pipeline in a fraction of the time and resources.

“We are always looking for new and better ways to analyze our huge data stores. SQream DB allows our Data Science team to generate the insights we need with a lean and performant addition to our existing Hadoop system.”
Team lead

Read more in the full case study

Global bank uses SQream DB for fraud detection - preview

Download customer success story: Global financial institution unlocks new insights