Case Study: edo

Overview:

edo provides a tailored, measurable and scalable card-linked offer platform to help marketers, merchants and banks increase customer loyalty. edo was not only growing its customer base at a tremendous rate, but they were developing more innovative ways in which to serve their customers. This growth and innovation led to a need for a more robust infrastructure. They consulted OpenBI, with their extensive experience with big data, BI and analytics, to determine the best technologies to meet their growing needs and to develop a plan to implement and transition to this new infrastructure. OpenBI recommended and ultimately deployed a Cloudera Hadoop and Vectorwise database hybrid data platform powered by Pentaho with Tableau and R for analytics.

Challenge:

“Using OpenBI to assist us shortened the time to production and the overall cost. They knew the market and were able to advise us on the best tools for our problem, while also delivering a high quality solution."

Tim Garnto, Vice President of Product Engineering, edo

With ever-increasing data volumes, edo needed an infrastructure that would be able to keep pace with the growing demands. Operational and analytical databases were beginning to show performance issues, reporting processes were slowing, and data build process times were increasing.

The Postgres database served edo well as a low-cost startup platform but was unable to support the heightened processing and analytic demands of a growing company. In addition, edo’s BI needs outpaced existing capabilities and edo required more dynamic analytical tools.

edo recognized the need to act proactively in order to continue to meet customer needs and to increase and improve their service offerings. They teamed with OpenBI to enact a solution.

Approach:

edo contracted OpenBI to recommend and execute a comprehensive big data and business intelligence solution. After developing a Roadmap, OpenBI collaborated with edo as part of an integrated team to implement the two-phase project: first enabling basic BI, then facilitating big data processing.

The team delivered the foundation of an expanded analytics infrastructure powered by a Hadoop cluster to ingest and process data. This gave edo the ability to collect information from their customers – marketers, merchants and banks – as well as from internal operational systems.

Pentaho big data capabilities were deployed to speed development and orchestrate processes. edo had successfully used Pentaho for existing ETL processing and liked its ability to offer visual development of in-cluster processing using Pentaho MapReduce. Pentaho also provided analytic data capabilities, allowing edo to explore and identify new opportunities for their customers.

Phase two focused on fully migrating to the Hadoop environment and subsequently retiring the existing data warehouse database, thus promoting more efficient customer information processing. The Pentaho MapReduce framework was used to transform raw data into usable information. Analysts now have direct access to a Hadoop-based data warehouse through Hive and Tableau Professional for analytical and modeling activities.

To enhance the information infrastructure, the team implemented scalable, more intelligent cardholder and merchant master data, in turn delivering orders of magnitude gain in targeting performance. Hbase was deployed to manage cardholder and merchant master data.

And to serve business users with rapid-fire analytic queries, the team deployed a high-performance analytic database, Vectorwise, from Actian. Columnar store Vectorwise was chosen for its consistently outstanding performance, scalability and low overall cost. With a combination of Vectorwise and Pentaho Analyzer, business users now have rapid access to data, supporting analytic queries they were unable to imagine before.

Results:

The migration to the Hadoop environment allowed edo to expand their analytics horizon and provided the ability to identify and explore new opportunities to benefit their customers. edo now has more robust analytical tools allowing more efficient and effective data analysis. Reports and processes run in a fraction of the previous time with less staff involvement. The scalable infrastructure facilitates the addition of new partners, supporting growth into the future.

 

Tags: 
Go to top