
Key highlights
Category
Financial Services/Retail
Location
Headquarters: Brazil
Solution highlights
- Modern Data Platform: Cloudera Enterprise
- Workloads: Data Warehouse
- Components: select from among Apache Sentry, Apache Impala, Cloudera Navigator
- Databases: Oracle Database Appliance
- BI & Analytics Tools: Qlik, SAS
Applications supported
- Customer 360
Data sources
- Credit card transactions
- Cardholder social channels
- Elo app beacon and mobile devices generating IoT data such as location
- Merchants
- Maps
- Weather
- Traffic
Impact
- Real-time, personalized offers drive a better customer experience
- New revenue stream from insights provided to merchants and banks
Big data scale
- 7 TB today, growing by 10 GB per day
With its real-time data warehousing on Cloudera, Elo provides personal, context-relevant communications to customers, while also offering data-driven insights to banks and merchants.
Cartao Elo, Brazil’s sole domestic credit card company, processes more than a million transactions per day, which represents 11 percent of the country's payment cards market share.
Challenge
Elo aims to enhance the consumer experience by delivering personalized, contextual offerings--before customers reach for their credit card. For instance, Elo wants to send restaurant patrons promotions for specific foods or beverages they’d be interested in during the one to two hour window that they’re in the restaurant. This requires a real-time analytics environment that can ingest data from sources such as social channels or merchant beacons, signaling information about consumers’ live behaviors and locations.
Elo’s legacy data management infrastructure couldn’t provide the performance or scalability to accommodate these new data sources. Business users were limited to canned, inflexible reports, and if users wanted to view different or additional data, they had to go through the data team which was time consuming. They’d need four to five DBAs just to satisfy all reporting requests.
Solution
Elo is improving the experience of both consumers and merchants with self-service, real-time analytics powered by a Cloudera data lake that combines transaction data with information from cardholders, merchants, social media, maps and weather.
The data lake supports three groups:
Operations, seeking to understand the volume of transactions and money spent with Elo cards.
Product, helping to understand where cardholders are spending money so they can improve products and services for customers.
Enablement, looking for opportunities to expand adoption of Elo cards across Brazil.
“Even our Chief Executive is looking at this data,” said Mauro Sartin, database manager for Cartao Elo.
Implementation
Elo set out to find a solution that could handle the scale and speed of Internet of Things (IoT) data, while accommodating the seasonality of its business and stringent security requirements for Payment Card Industry (PCI) compliance.
With flexibility to support both on-premises and cloud-based deployments, Cloudera makes it easy for Elo to expand and shrink its cluster as demands change--for instance, supporting transaction peaks around holidays such as Christmas, Mother’s Day and Father’s Day.
Cloudera Navigator provides data encryption, columnar encryption and access controls for data and users, simplifying the path to PCI compliance. This is critically important for a company dealing with international credit card data; Elo’s partnership with Discover allows consumers to buy items in the U.S., but without PCI compliance, that partnership would not be possible.
Results
By capturing non-transaction-based data from social channels, merchant beacons, and other new sources, Elo is equipped to provide personal, context-relevant communications to customers, while also offering data-driven insights to banks and merchants. This represents a new revenue stream for Elo.
“The bank had more information until this point than Elo. Now, we have more information than the bank and even the merchants,” said Sartin. “We are starting to provide insights about how the holder behaves and the banks will pay us money for that. With Cloudera, we can provide the best user experience because we process data fast enough to reach the customer before they use our cards.”